{"as_of":"2026-08-08T04:28:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:21ab2fb93085bd02ad5dce3aed4d0d7040cb633452b1ccb1241156c6a4191c37","coverage":[{"denominator":66,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":66,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-07T10:52:49.765795Z","state":"measured"},{"denominator":66,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":66,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-07T06:34:17.273281+00:00","state":"measured"},{"denominator":0,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[],"links":{"evidence":"/evidence","html":"/paper/2506.04205/citation-record","integrity":"/paper/2506.04205/integrity","json":"/paper/2506.04205/citation-record.json","paper":"/paper/2506.04205"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.561414Z","title":"Chain-of-thought prompting elicits reasoning in large language models.Advances in neural information processing systems, 35:24824–24837, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.561414Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:0d97397f8dcd9da3f5248e6357d85ad45144ec4dce039f4015d6cc63ed8241a1","observation_id":"7553e068-703d-4ea9-b922-0ba1f3cb8889","resolution":{"observed_at":"2026-08-07T10:52:49.561414Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.565459Z","title":"Let’s verify step by step","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.565459Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:93b8079af3dad4f211225426d24c180369dd7a8dfde9756b17d1e722e966bc85","observation_id":"50f87e97-a253-44ae-ac0b-44a8b78d8e6a","resolution":{"observed_at":"2026-08-07T10:52:49.565459Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.576451Z","title":"Gpqa: A graduate-level google-proof q&a benchmark","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.576451Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:5305a9a3dcc7f90f8d39376db29cc7b81edf2c7077df1539040eff933be7b25d","observation_id":"1fd0f809-d993-4ac4-8ea0-e48496cc9145","resolution":{"observed_at":"2026-08-07T10:52:49.576451Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2501.12599","last_updated":"2025-06-03T02:14:54Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2025-01-22T02:48:14Z","title":"Kimi k1.5: Scaling Reinforcement Learning with LLMs","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.12599","snapshot_observed_at":"2026-08-07T10:52:49.579885Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.579885Z"},"links":{"cited_paper":"/paper/2501.12599","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:493fc2b71a5c0f84d414a4c804e64849c415afa2a990b935a7967fd5b9d80e13","observation_id":"ddc22cce-f9d7-46fc-b52d-4d52b1e54917","resolution":{"observed_at":"2026-08-07T10:52:49.579885Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.583827Z","title":"Open r1: A fully open reproduction of deepseek-r1, January 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.583827Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:9b24a3dd1f758d53a1bf3295b0d010ead837f3de4a0340e37fb151143163bdb7","observation_id":"bbecbb67-53bf-4cad-836a-d7baaeee0af9","resolution":{"observed_at":"2026-08-07T10:52:49.583827Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.586794Z","title":"Open Thoughts","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.586794Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:573a8856fc3a5b79a0484a92859171e6e0f8578b284944e055a295e2065343f1","observation_id":"195b63ab-3ea6-489d-aa90-7f25ddbe0d50","resolution":{"observed_at":"2026-08-07T10:52:49.586794Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2501.19393","last_updated":"2025-03-01T06:07:39Z","snapshot_observed_at":"2026-07-06T20:29:11.710285Z","submitted_at":"2025-01-31T18:48:08Z","title":"s1: Simple test-time scaling","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.19393","snapshot_observed_at":"2026-08-07T10:52:49.589581Z","title":"s1: Simple test-time scaling","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.589581Z"},"links":{"cited_paper":"/paper/2501.19393","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:b4ff3c65183189c5066991de0156b0b806868cf8aa29390162e1caa76d21490f","observation_id":"af91db63-914f-4f15-846c-b4f0d20c1ef8","resolution":{"observed_at":"2026-08-07T10:52:49.589581Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2502.03387","last_updated":"2025-07-29T16:23:02Z","snapshot_observed_at":"2026-08-08T04:13:22.884923Z","submitted_at":"2025-02-05T17:23:45Z","title":"LIMO: Less is More for Reasoning","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2502.03387","snapshot_observed_at":"2026-08-07T10:52:49.592653Z","title":"Limo: Less is more for reasoning","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.592653Z"},"links":{"cited_paper":"/paper/2502.03387","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:4a05e94e46c228ce115f18d5a08733353d1c3fcebebb715c9cf0a83fa3389e03","observation_id":"7777abc1-fd26-42ad-8b8f-9d2b9a8b2d3b","resolution":{"observed_at":"2026-08-07T10:52:49.592653Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2412.21187","last_updated":"2025-02-01T07:57:37Z","snapshot_observed_at":"2026-08-01T16:43:44.704797Z","submitted_at":"2024-12-30T18:55:12Z","title":"Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.21187","snapshot_observed_at":"2026-08-07T10:52:49.595462Z","title":"Do not think that much for 2+ 3=? on the overthinking of o1-like llms","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.595462Z"},"links":{"cited_paper":"/paper/2412.21187","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:eb1c7fae897af977d604b4b3a9b590ed212f6b93454ce2f61e2d1056d45be122","observation_id":"ee49e7df-3b1f-41f7-8c25-5f8595217cf2","resolution":{"observed_at":"2026-08-07T10:52:49.595462Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2501.18585","last_updated":"2025-02-18T16:51:53Z","snapshot_observed_at":"2026-07-06T20:28:33.378335Z","submitted_at":"2025-01-30T18:58:18Z","title":"Thoughts Are All Over the Place: On the Underthinking of o1-Like LLMs","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.18585","snapshot_observed_at":"2026-08-07T10:52:49.599042Z","title":"Thoughts are all over the place: On the underthinking of o1-like llms","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.599042Z"},"links":{"cited_paper":"/paper/2501.18585","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:c605152870e287ccfeed1ab3e09f997416deff3aed2812ca9298370ce7e7f0d3","observation_id":"02cac6bf-36c2-461b-b13f-783b4d571ac2","resolution":{"observed_at":"2026-08-07T10:52:49.599042Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2211.09066","last_updated":"2022-11-15T06:12:28Z","snapshot_observed_at":"2026-08-06T04:36:04.765232Z","submitted_at":"2022-11-15T06:12:28Z","title":"Teaching Algorithmic Reasoning via In-context Learning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2211.09066","snapshot_observed_at":"2026-08-07T10:52:49.602032Z","title":"Teaching algorithmic reasoning via in-context learning","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.602032Z"},"links":{"cited_paper":"/paper/2211.09066","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:4e9f9a083bf9208bbe658bab12957d8898f900974682cfd4674afaf27c67fd30","observation_id":"3e091742-8cc5-4d2d-b6a8-15fa6949332e","resolution":{"observed_at":"2026-08-07T10:52:49.602032Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.622210Z","title":"Distilling reasoning capabilities into smaller language models","venue":null,"work_id":"9bb4669a-4bc2-4d74-9985-4d0d8cfab7eb","year":2023},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.607192Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:428c16ea1f8ee4c1565ce1089816cfe245978ee6944b3c31c08e9d92dac95785","observation_id":"4d4ee36d-7bc7-4e4f-8264-c855b315397a","resolution":{"observed_at":"2026-08-07T10:52:50.625911Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.613088Z","title":"Specializing smaller language models towards multi-step reasoning","venue":null,"work_id":"e87aab88-39de-4c77-adff-747747e4c3ae","year":2023},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.610873Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:3fdc18aac3bc51bd2f46d0cfaa310b5db863409a755613f0e649a48c73780cbc","observation_id":"e5701677-c9d1-41f6-ab14-165bb5d1c76f","resolution":{"observed_at":"2026-08-07T10:52:50.616204Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.604232Z","title":"Generalthought-195k","venue":null,"work_id":"f2418eae-1a86-4f20-965c-00515b450ef8","year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.613898Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:cb92638c6284b921d22f7f541ad677d6c64c405e386ebcabec2bfed0fa124100","observation_id":"1c158de0-7333-4b42-a14a-49b51b670ad6","resolution":{"observed_at":"2026-08-07T10:52:50.607001Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.594877Z","title":"Sky-t1: Train your own o1 preview model within $450","venue":null,"work_id":"1f3a41a4-345d-4c79-96a7-c4ea113525d1","year":null},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.616668Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:0a8d328b8d58a9e03ddefa1748b0bcc0f8bbead676bde6b0798480519961f198","observation_id":"2f37a542-5d93-4922-afa8-2dbfaf206b10","resolution":{"observed_at":"2026-08-07T10:52:50.597853Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.622263Z","title":"Bespoke-stratos: The unreasonable effectiveness of reasoning distillation","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.622263Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:67b50725d26cdfc90fb0190882d4a186be187f87909148fc6dfc38e40041e0ca","observation_id":"93843c2a-49ba-48e3-9d0c-781d19a30cab","resolution":{"observed_at":"2026-08-07T10:52:49.622263Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.572537Z","title":"Re-distilling smaller deepseek r1 models for better performance, January 2025","venue":null,"work_id":"0a324acb-c295-40cd-8d6f-5a17186cec07","year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.625119Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:2bc19593b1a691ae02dc705b7176b40e7ec0965396edf0f969f93ea7957ad5b3","observation_id":"9e8cc1a5-c084-4221-ac3d-a6e3ebf81d9f","resolution":{"observed_at":"2026-08-07T10:52:50.575878Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.627835Z","title":"Small models struggle to learn from strong reasoners","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.627835Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:d4c8feeb49f6e39ff1197119af15a8093c3e5444d585d780039b1e9baf867453","observation_id":"b67a320c-a385-418c-9cb2-ac8cc0a2f360","resolution":{"observed_at":"2026-08-07T10:52:49.627835Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2501.11284","last_updated":"2025-01-20T05:44:01Z","snapshot_observed_at":"2026-08-06T13:30:45.600197Z","submitted_at":"2025-01-20T05:44:01Z","title":"RedStar: Does Scaling Long-CoT Data Unlock Better Slow-Reasoning Systems?","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.11284","snapshot_observed_at":"2026-08-07T10:52:49.630323Z","title":"Redstar: Does scaling long-cot data unlock better slow-reasoning systems? arXiv preprint arXiv:2501.11284, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.630323Z"},"links":{"cited_paper":"/paper/2501.11284","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:3b87b5d45f631f30819033bfc60eb2a94d37869e1ae90b0e0a5c14d4f517c2cf","observation_id":"292b9609-9766-4900-a4d7-9b02281e81cf","resolution":{"observed_at":"2026-08-07T10:52:49.630323Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2503.02875","last_updated":"2025-03-04T18:56:03Z","snapshot_observed_at":"2026-08-07T17:29:34.418079Z","submitted_at":"2025-03-04T18:56:03Z","title":"The First Few Tokens Are All You Need: An Efficient and Effective Unsupervised Prefix Fine-Tuning Method for Reasoning Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2503.02875","snapshot_observed_at":"2026-08-07T10:52:49.633523Z","title":"The first few tokens are all you need: An efficient and effective unsupervised prefix fine-tuning method for reasoning models","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.633523Z"},"links":{"cited_paper":"/paper/2503.02875","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:1c30f31182f01bccdf2bd90838b6f32bfdc2fdd622155a5f8c8fcd5f83e583f9","observation_id":"fa8cbb43-9442-421d-a8db-d66217ecef7d","resolution":{"observed_at":"2026-08-07T10:52:49.633523Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2201.11903","last_updated":"2023-01-10T23:07:57Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2022-01-28T02:33:07Z","title":"Chain-of-Thought Prompting Elicits Reasoning in Large Language Models","version":6},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2201.11903","snapshot_observed_at":"2026-08-07T10:52:49.636416Z","title":"Chain-of-thought prompting elicits reasoning in large language models, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.636416Z"},"links":{"cited_paper":"/paper/2201.11903","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:c489766460be85917f162f1da4634e7bda640df0af84dd24be815828354bf96d","observation_id":"2862761c-9bea-4600-9376-da6e85da8cd7","resolution":{"observed_at":"2026-08-07T10:52:49.636416Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.00724","last_updated":"2025-03-03T07:53:32Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-08-01T17:16:04Z","title":"Inference Scaling Laws: An Empirical Analysis of Compute-Optimal Inference for Problem-Solving with Language Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.00724","snapshot_observed_at":"2026-08-07T10:52:49.639054Z","title":"Inference scaling laws: An empirical analysis of compute-optimal inference for problem-solving with language models, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.639054Z"},"links":{"cited_paper":"/paper/2408.00724","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:bd4a18c6c4fb62b12105d115321b222a1b760ee910dfc057cd40d393dd4d2054","observation_id":"ab17a2dc-1f40-4848-b9d5-d2a57f276ad7","resolution":{"observed_at":"2026-08-07T10:52:49.639054Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2501.12948","last_updated":"2026-01-04T03:57:36Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2025-01-22T15:19:35Z","title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.12948","snapshot_observed_at":"2026-08-07T10:52:49.642382Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.642382Z"},"links":{"cited_paper":"/paper/2501.12948","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:cc8c81c8af9103e6ce1220f14a1012ee39b02497401986d136f7508ed1ec1321","observation_id":"ec81d348-18a7-4294-80fd-43521d69084a","resolution":{"observed_at":"2026-08-07T10:52:49.642382Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.03314","last_updated":"2024-08-06T17:35:05Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-08-06T17:35:05Z","title":"Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.03314","snapshot_observed_at":"2026-08-07T10:52:49.645112Z","title":"Scaling llm test-time compute optimally can be more effective than scaling model parameters, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.645112Z"},"links":{"cited_paper":"/paper/2408.03314","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:6ed47b2c0c38fca769fa869181c080e79780df7d7d1421fa26596f7d7b92a6a5","observation_id":"ac419b76-f34b-485f-9ab3-204195e48e25","resolution":{"observed_at":"2026-08-07T10:52:49.645112Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2203.11171","last_updated":"2023-03-07T17:57:37Z","snapshot_observed_at":"2026-07-06T12:50:22.773056Z","submitted_at":"2022-03-21T17:48:52Z","title":"Self-Consistency Improves Chain of Thought Reasoning in Language Models","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2203.11171","snapshot_observed_at":"2026-08-07T10:52:49.647879Z","title":"Self-consistency improves chain of thought reasoning in language models, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.647879Z"},"links":{"cited_paper":"/paper/2203.11171","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:419a26af7b52e24cc70ad4cd8d0388ed012122c1292c23bf4dc2f4db9fe4e7b3","observation_id":"2e8fc576-de5a-4077-a0db-c7d85e05a2dd","resolution":{"observed_at":"2026-08-07T10:52:49.647879Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.11860","last_updated":"2023-11-16T16:47:05Z","snapshot_observed_at":"2026-08-04T02:42:35.223564Z","submitted_at":"2023-05-19T17:49:25Z","title":"Let's Sample Step by Step: Adaptive-Consistency for Efficient Reasoning and Coding with LLMs","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.11860","snapshot_observed_at":"2026-08-07T10:52:49.650755Z","title":"Let’s sample step by step: Adaptive-consistency for efficient reasoning and coding with llms","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.650755Z"},"links":{"cited_paper":"/paper/2305.11860","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:fc8f507179bc52a1dc7ad76fd5d6cf8a389c5b95d43475138489a118ab221beb","observation_id":"24fde902-077c-4fba-bc4f-1895e8a99332","resolution":{"observed_at":"2026-08-07T10:52:49.650755Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2407.21787","last_updated":"2024-12-30T19:03:24Z","snapshot_observed_at":"2026-07-06T18:55:11.576666Z","submitted_at":"2024-07-31T17:57:25Z","title":"Large Language Monkeys: Scaling Inference Compute with Repeated Sampling","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.21787","snapshot_observed_at":"2026-08-07T10:52:49.653723Z","title":"Large language monkeys: Scaling inference compute with repeated sampling","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.653723Z"},"links":{"cited_paper":"/paper/2407.21787","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:8443dd2017830044b1ac69d909dbcb525baf3694a9d8384c2ae8788117f8fbdc","observation_id":"22c3ab18-3fc2-4c39-9110-3702a75f7411","resolution":{"observed_at":"2026-08-07T10:52:49.653723Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.10601","last_updated":"2023-12-03T22:50:35Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-05-17T23:16:17Z","title":"Tree of Thoughts: Deliberate Problem Solving with Large Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.10601","snapshot_observed_at":"2026-08-07T10:52:49.657309Z","title":"Griffiths, Yuan Cao, and Karthik Narasimhan","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.657309Z"},"links":{"cited_paper":"/paper/2305.10601","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:3f5e3a9fd296342a4a4603c1e40b118267e2edbb75c48258abd43da5bf6aeab0","observation_id":"33260b3a-5eda-464a-be41-d96f88e50944","resolution":{"observed_at":"2026-08-07T10:52:49.657309Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.08152","last_updated":"2024-08-15T13:40:03Z","snapshot_observed_at":"2026-07-06T19:01:06.124684Z","submitted_at":"2024-08-15T13:40:03Z","title":"DeepSeek-Prover-V1.5: Harnessing Proof Assistant Feedback for Reinforcement Learning and Monte-Carlo Tree Search","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.08152","snapshot_observed_at":"2026-08-07T10:52:49.660122Z","title":null,"venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.660122Z"},"links":{"cited_paper":"/paper/2408.08152","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:256444a49c91931cd382e90de0adef2b8e13c70d09c5394e4e9d4a76ac57fb34","observation_id":"0fffaa7c-c3cf-4c90-a600-e3b2b07038c1","resolution":{"observed_at":"2026-08-07T10:52:49.660122Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.563589Z","title":"Generat- ing sequences by learning to self-correct","venue":null,"work_id":"aede6636-65f5-4329-a56c-9dcc6228781a","year":null},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.663070Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:fd31ee817c79ae844ff0eb2330ce4d2d6fc9f0a6a284fd3185e852dd280e143f","observation_id":"3d3d6509-1b72-4bc0-9dbe-98e0cdd70a1c","resolution":{"observed_at":"2026-08-07T10:52:50.566674Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2303.17651","last_updated":"2023-05-25T19:13:47Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-03-30T18:30:01Z","title":"Self-Refine: Iterative Refinement with Self-Feedback","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2303.17651","snapshot_observed_at":"2026-08-07T10:52:49.669276Z","title":"Self-refine: Iterative refinement with self-feedback, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.669276Z"},"links":{"cited_paper":"/paper/2303.17651","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:aee409ce1a163e9d45074cc61dbd7c728625e3eac9872c0b00d2cc726a1551fd","observation_id":"954b2356-6baf-43bb-8a3a-55e8fb620164","resolution":{"observed_at":"2026-08-07T10:52:49.669276Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.544691Z","title":"From decoding to meta-generation: Inference-time algorithms for large language models,","venue":null,"work_id":"9acdf935-621d-46d6-8f76-b2cab4103037","year":null},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.672528Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:4b1eece0a429e17b1e2786444fafe1acb7c8a4c0184b16fe21cc450d98586ec8","observation_id":"0ae6e127-f190-44a8-a724-e4e65beda43c","resolution":{"observed_at":"2026-08-07T10:52:50.547842Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2502.06703","last_updated":"2025-02-10T17:30:23Z","snapshot_observed_at":"2026-08-05T04:32:32.994993Z","submitted_at":"2025-02-10T17:30:23Z","title":"Can 1B LLM Surpass 405B LLM? Rethinking Compute-Optimal Test-Time Scaling","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2502.06703","snapshot_observed_at":"2026-08-07T10:52:49.678833Z","title":"Can 1b llm surpass 405b llm? rethinking compute-optimal test-time scaling","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.678833Z"},"links":{"cited_paper":"/paper/2502.06703","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:2e3018873cb78b65a7054784c579e4d0fe32b9f4110692106ddbe314859c9daa","observation_id":"df98069d-c01a-4be8-a355-bf5788b300b8","resolution":{"observed_at":"2026-08-07T10:52:49.678833Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2503.04697","last_updated":"2025-10-03T01:55:58Z","snapshot_observed_at":"2026-08-06T08:53:09.095000Z","submitted_at":"2025-03-06T18:43:29Z","title":"L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2503.04697","snapshot_observed_at":"2026-08-07T10:52:49.682029Z","title":"L1: Controlling how long a reasoning model thinks with reinforcement learning","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.682029Z"},"links":{"cited_paper":"/paper/2503.04697","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:e6629af8d172345867be53d611e46aabf8cd0da2b2e951084beaa455da9d9c35","observation_id":"6dad952b-f63e-466c-ad18-00dbd4330eb0","resolution":{"observed_at":"2026-08-07T10:52:49.682029Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2501.12570","last_updated":"2025-01-29T03:11:03Z","snapshot_observed_at":"2026-08-08T03:22:47.699927Z","submitted_at":"2025-01-22T01:35:11Z","title":"O1-Pruner: Length-Harmonizing Fine-Tuning for O1-Like Reasoning Pruning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.12570","snapshot_observed_at":"2026-08-07T10:52:49.685088Z","title":"O1-pruner: Length-harmonizing fine-tuning for o1-like reasoning pruning","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.685088Z"},"links":{"cited_paper":"/paper/2501.12570","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:3af0d3ea92a02cc571f233ac90a3b8a36d78558d846708df3d6fb1e5e022babd","observation_id":"85b05d1e-a5fb-488f-b15f-ce1a24ef4d6b","resolution":{"observed_at":"2026-08-07T10:52:49.685088Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.687820Z","title":"Tokenskip: Controllable chain-of-thought compression in llms","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.687820Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:a7ff7ffe665f627e8be6e6a769e361e011481e190ba56a64f1549ffabec0f976","observation_id":"e415a5cc-c3c6-439f-93cb-5fbc81e808a2","resolution":{"observed_at":"2026-08-07T10:52:49.687820Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.690806Z","title":"Lightthinker: Thinking step-by-step compression","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.690806Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:a7f1fa1178bbe0ee3a5b301171e383de49d3c2561381436129e777920922d9a6","observation_id":"fd3a15e7-32cf-4d3b-aed4-a13496f0d60b","resolution":{"observed_at":"2026-08-07T10:52:49.690806Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.535270Z","title":"Similar: Submodular information measures based active learning in realistic scenarios","venue":null,"work_id":"f569c41e-09ab-47ea-9f62-d203bff1e7c5","year":2021},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.694028Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:eacc23680ee412048dcda28636e2ab00b981adf229f2a39bf24ed161ae1506d2","observation_id":"7717239d-8a7a-4d08-97f6-dd607b20b9f7","resolution":{"observed_at":"2026-08-07T10:52:50.538230Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2103.00123","last_updated":"2021-06-11T22:08:29Z","snapshot_observed_at":"2026-07-06T10:45:06.227629Z","submitted_at":"2021-02-27T04:09:32Z","title":"GRAD-MATCH: Gradient Matching based Data Subset Selection for Efficient Deep Model Training","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2103.00123","snapshot_observed_at":"2026-08-07T10:52:49.697076Z","title":"Grad-match: A gradient matching based data subset selection for efficient learning","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.697076Z"},"links":{"cited_paper":"/paper/2103.00123","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:bdf74b2c6dd6d7779f7c890a375888745c52442869a3fabeccb3ca86e86503f8","observation_id":"b0d60cd4-46fd-45a8-a180-5172c1066f19","resolution":{"observed_at":"2026-08-07T10:52:49.697076Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2107.06499","last_updated":"2022-03-24T19:29:45Z","snapshot_observed_at":"2026-08-05T18:30:13.997811Z","submitted_at":"2021-07-14T06:06:52Z","title":"Deduplicating Training Data Makes Language Models Better","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2107.06499","snapshot_observed_at":"2026-08-07T10:52:49.699909Z","title":"Deduplicating training data makes language models better","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.699909Z"},"links":{"cited_paper":"/paper/2107.06499","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:4da0a474b5271ccc1baa221ec2efd285532b042938ee784c29116fabe56886f2","observation_id":"0f0b9125-7aaa-41fe-9713-d48a9a0e84c2","resolution":{"observed_at":"2026-08-07T10:52:49.699909Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.525855Z","title":"Dataset pruning for resource-constrained spoofed audio detection","venue":null,"work_id":"bc0f3328-9f65-4773-b0e7-ada7c2fb3634","year":2022},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.703093Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:2adfec37fd672036bc8a3ca0547e6deb4fe887a12ad72ad5de432b06c7bf6634","observation_id":"8d7a7f66-6f0f-4f61-b53f-1ce86d42f75c","resolution":{"observed_at":"2026-08-07T10:52:50.528742Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.517031Z","title":"Contextual diversity for active learning","venue":null,"work_id":"08556650-5d73-43f3-969b-7bd39676f9f0","year":2020},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.705742Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:724929183d6fc28e482416de0fb08855b3f0fdeb7b4f3f474da1e59854fd074a","observation_id":"1e0e364c-f685-41d0-9fa9-439fd248fe3e","resolution":{"observed_at":"2026-08-07T10:52:50.520002Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1906.11829","last_updated":"2020-10-27T00:52:20Z","snapshot_observed_at":"2026-07-06T08:03:24.054624Z","submitted_at":"2019-06-26T23:01:47Z","title":"Selection via Proxy: Efficient Data Selection for Deep Learning","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1906.11829","snapshot_observed_at":"2026-08-07T10:52:49.708522Z","title":"Selection via proxy: Efficient data selection for deep learning","venue":null,"work_id":null,"year":1906},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.708522Z"},"links":{"cited_paper":"/paper/1906.11829","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:da3a936f021ee925fd95dd4e02c58309db24482d5ac26786f91e42ee6148e750","observation_id":"1f85362d-36e4-43f4-aac3-915b1cff94a6","resolution":{"observed_at":"2026-08-07T10:52:49.708522Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.507994Z","title":"Active learning is a strong baseline for data subset selection","venue":null,"work_id":"1fa0dba5-99e6-4f21-b536-edd108152b01","year":2022},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.711400Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:5f4c7de4b24fc3982aea60de4134e2aa53877c53e3d888da28421cbbec1113f0","observation_id":"87e23ff1-7ecd-480f-b882-04e06473a068","resolution":{"observed_at":"2026-08-07T10:52:50.511019Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.498562Z","title":"Coresets for data-efficient training of machine learning models","venue":null,"work_id":"1c596eb4-c4c5-43dd-a219-9fd0426c8077","year":2020},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.714159Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:90f85d6f8386dd5046bea8f8ed7ed932c612d709d16e6f850b087d691642c0c0","observation_id":"627d5ee4-0d74-47b6-a819-e74f223fa45b","resolution":{"observed_at":"2026-08-07T10:52:50.501816Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1812.05159","last_updated":"2019-11-15T17:08:30Z","snapshot_observed_at":"2026-08-01T06:33:25.731787Z","submitted_at":"2018-12-12T21:24:15Z","title":"An Empirical Study of Example Forgetting during Deep Neural Network Learning","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1812.05159","snapshot_observed_at":"2026-08-07T10:52:49.717004Z","title":"An empirical study of example forgetting during deep neural network learning","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.717004Z"},"links":{"cited_paper":"/paper/1812.05159","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:18f768f00e862e3b5d3a3fd60261b03ce10446888888a16677047d4f46a2c8c8","observation_id":"8866c768-dc99-4b75-8919-3d6ccde2859e","resolution":{"observed_at":"2026-08-07T10:52:49.717004Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.719767Z","title":"Deep learning on a data diet: Finding important examples early in training","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.719767Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:d5b971286be92250b57585ee9133fc50a649f0738301e6db70fa868c40ab0119","observation_id":"2f877161-bec9-431a-9ce1-017f6992099d","resolution":{"observed_at":"2026-08-07T10:52:49.719767Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2312.08901","last_updated":"2024-02-15T05:42:15Z","snapshot_observed_at":"2026-07-06T17:01:40.744034Z","submitted_at":"2023-12-14T13:03:13Z","title":"Fewer is More: Boosting LLM Reasoning with Reinforced Context Pruning","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2312.08901","snapshot_observed_at":"2026-08-07T10:52:49.722603Z","title":"Fewer is more: Boosting llm reasoning with reinforced context pruning","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.722603Z"},"links":{"cited_paper":"/paper/2312.08901","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:c4c2f4efd93a1b842575d7711c5abd3fe717589ea8fe10195696c0761246568d","observation_id":"c295cae3-eb0a-489c-8b56-786356d1ea2b","resolution":{"observed_at":"2026-08-07T10:52:49.722603Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.483067Z","title":"Staff: Specu- lative coreset selection for task-specific fine-tuning","venue":null,"work_id":"c08d3384-e62d-47bd-a243-c52abc4495d5","year":null},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.725641Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:f48a81e6dc37bc4f0feca65eaeab361e97507aec46830c981d5c527c835375bc","observation_id":"4bace416-87e2-43e9-9152-520cdf5dcb52","resolution":{"observed_at":"2026-08-07T10:52:50.486551Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2402.04333","last_updated":"2024-06-13T03:42:02Z","snapshot_observed_at":"2026-08-07T17:47:02.067367Z","submitted_at":"2024-02-06T19:18:04Z","title":"LESS: Selecting Influential Data for Targeted Instruction Tuning","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.04333","snapshot_observed_at":"2026-08-07T10:52:49.728829Z","title":"Less: Selecting influential data for targeted instruction tuning","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.728829Z"},"links":{"cited_paper":"/paper/2402.04333","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:b3e33f2b38613ee9920fd98ff0830b32a970caf1c9f8fa869eba8dc96388a465","observation_id":"c2e64d93-be90-45dc-b914-c03dbf45b9ce","resolution":{"observed_at":"2026-08-07T10:52:49.728829Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.473824Z","title":"Lima: Less is more for alignment","venue":null,"work_id":"a7b7ead4-c036-4f57-a4e1-fc4647ae78bb","year":2023},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.731697Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:4b59236276e7ea95f1c508eb7ebb9fb3d13ed16b4f2154b979d2666be10da6f7","observation_id":"dce7c1fa-8955-4eca-b5ed-8c3abc88b223","resolution":{"observed_at":"2026-08-07T10:52:50.476706Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2412.16720","last_updated":"2026-04-30T02:46:40Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-12-21T18:04:31Z","title":"OpenAI o1 System Card","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.16720","snapshot_observed_at":"2026-08-07T10:52:49.734344Z","title":"Openai o1 system card","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":54,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.734344Z"},"links":{"cited_paper":"/paper/2412.16720","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:2b2c0b4447304be4e53d81ce1c364b8ff7f651dfe7acb68af71cf17055aad7dc","observation_id":"5611cfc9-3ddf-4b9b-823c-ac5fb56c7e39","resolution":{"observed_at":"2026-08-07T10:52:49.734344Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2412.09413","last_updated":"2024-12-22T10:44:13Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-12-12T16:20:36Z","title":"Imitate, Explore, and Self-Improve: A Reproduction Report on Slow-thinking Reasoning Systems","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.09413","snapshot_observed_at":"2026-08-07T10:52:49.737129Z","title":"Imitate, explore, and self-improve: A reproduction report on slow-thinking reasoning systems","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":55,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.737129Z"},"links":{"cited_paper":"/paper/2412.09413","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:9d06e4f37830433780e2c1cf455ff347372ac83d061dd436615acf835407124d","observation_id":"bde1899b-7687-464e-a0de-9d3016f4d899","resolution":{"observed_at":"2026-08-07T10:52:49.737129Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2403.12968","last_updated":"2024-08-12T04:48:11Z","snapshot_observed_at":"2026-08-05T09:53:46.200081Z","submitted_at":"2024-03-19T17:59:56Z","title":"LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2403.12968","snapshot_observed_at":"2026-08-07T10:52:49.739784Z","title":"Llmlingua-2: Data distillation for efficient and faithful task-agnostic prompt compression","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":56,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.739784Z"},"links":{"cited_paper":"/paper/2403.12968","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:159fc76bc7b6a06b9f521586e6774875a6d929d13d3d6d6541106c1a0de11e03","observation_id":"f6d45ce7-ba34-4679-bd80-49b8aa172fb1","resolution":{"observed_at":"2026-08-07T10:52:49.739784Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1705.04146","last_updated":"2017-10-23T16:45:03Z","snapshot_observed_at":"2026-08-02T16:52:38.363314Z","submitted_at":"2017-05-11T13:04:47Z","title":"Program Induction by Rationale Generation : Learning to Solve and Explain Algebraic Word Problems","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1705.04146","snapshot_observed_at":"2026-08-07T10:52:49.742589Z","title":"Program induction by rationale generation: Learning to solve and explain algebraic word problems","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":57,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.742589Z"},"links":{"cited_paper":"/paper/1705.04146","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:4a9e6ea562099f504021c54e38e7415b446797d0f6d3ac46443e77a725cd1c3a","observation_id":"59ca9d86-e389-49b8-8d27-c2c2cf0496b6","resolution":{"observed_at":"2026-08-07T10:52:49.742589Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.745278Z","title":"Land- scape of thoughts: Visualizing the reasoning process of large language models","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":58,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.745278Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:9a620584435e694fecc6084a075315122c5ab37d25b3c0b001dacf7f076cb2b1","observation_id":"7e4f9db4-eb03-49f9-93fe-6679630c599b","resolution":{"observed_at":"2026-08-07T10:52:49.745278Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.748222Z","title":"Estimating mutual information","venue":null,"work_id":null,"year":2004},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":59,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.748222Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:ce4bf5003673305563a206956e72c28ef627ce38ad3c4866016b9f56b9c02d72","observation_id":"e8525c51-7dc8-4bcd-bfaa-57e1b822fb37","resolution":{"observed_at":"2026-08-07T10:52:49.748222Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.750747Z","title":"Pointer sentinel mixture models, 2016","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":60,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.750747Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:9a8322bf8c180e02255aaa6aac8e0e35c6ac127198c15e847f656dfb4ea5e911","observation_id":"3fa628d8-a111-49da-bcc2-bb35d46ca8b5","resolution":{"observed_at":"2026-08-07T10:52:49.750747Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.451005Z","title":"Math-verify: Math verification library, 2024","venue":null,"work_id":"22de44ea-5cf0-4c9d-8d4b-17efe407377b","year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":61,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.753709Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:5a8c1ec0de2ab0a873e63f40b6058801ac5fa1e6dfeccb4a23e74f739f61e90b","observation_id":"f255497d-6f64-47a9-83c7-2e0d11c5270c","resolution":{"observed_at":"2026-08-07T10:52:50.454743Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2409.12122","last_updated":"2024-09-18T16:45:37Z","snapshot_observed_at":"2026-07-06T19:17:41.512834Z","submitted_at":"2024-09-18T16:45:37Z","title":"Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvement","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2409.12122","snapshot_observed_at":"2026-08-07T10:52:49.756502Z","title":null,"venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":62,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.756502Z"},"links":{"cited_paper":"/paper/2409.12122","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:8a5e7796a88580fb3ac579c9ddacc6945e2062295fe69bd86a3e24a0434aa42a","observation_id":"e2818539-ad52-4a83-8805-e7d8b4f331cb","resolution":{"observed_at":"2026-08-07T10:52:49.756502Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2412.15115","last_updated":"2025-01-03T02:18:21Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-12-19T17:56:09Z","title":"Qwen2.5 Technical Report","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.15115","snapshot_observed_at":"2026-08-07T10:52:49.759907Z","title":null,"venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":63,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.759907Z"},"links":{"cited_paper":"/paper/2412.15115","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:084e6652d70c7bd7028bac1641e9de382aadb46e8f81f1bccd0f2489cfeab45b","observation_id":"a818a938-fc08-46d3-a528-d6681a357f7c","resolution":{"observed_at":"2026-08-07T10:52:49.759907Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2407.21783","last_updated":"2024-11-23T23:27:33Z","snapshot_observed_at":"2026-07-06T18:55:11.576666Z","submitted_at":"2024-07-31T17:54:27Z","title":"The Llama 3 Herd of Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.21783","snapshot_observed_at":"2026-08-07T10:52:49.762772Z","title":"The llama 3 herd of models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":64,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.762772Z"},"links":{"cited_paper":"/paper/2407.21783","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:272e25fddf3d824255e21e910eb5ab7307e072e4799122ce125a3db4e16e67fc","observation_id":"cb333f73-2947-4d11-b394-2c2cac17f800","resolution":{"observed_at":"2026-08-07T10:52:49.762772Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.440443Z","title":"Aime problems and solutions","venue":null,"work_id":"b15c7067-159f-4201-97ba-db75625d5106","year":null},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":65,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.765795Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:8b7acb7ac4d9d5ab16e617825f33ceaec74b5eccc104b294e385a3af8c9c49b2","observation_id":"7b486ab1-7932-4020-bbfd-c639a059e38a","resolution":{"observed_at":"2026-08-07T10:52:50.444190Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:50.554142Z","title":null,"venue":null,"work_id":"5600ad4d-9ac4-49a2-8b50-44a3e7ace6ee","year":null},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":2023,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.666542Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:d1b6889c7c602574dbb8228478d0456c4a8391d53ca0a4e2e677cd57218b1668","observation_id":"fe61110a-0fb6-48b3-b523-7bf880605ae6","resolution":{"observed_at":"2026-08-07T10:52:50.557367Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2406.16838","last_updated":"2024-11-20T17:57:26Z","snapshot_observed_at":"2026-07-06T18:36:12.298104Z","submitted_at":"2024-06-24T17:45:59Z","title":"From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.16838","snapshot_observed_at":"2026-08-07T10:52:49.675555Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":2024,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.675555Z"},"links":{"cited_paper":"/paper/2406.16838","citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:3d14484984a536bfc84ffc40e5278531b4cafc0c10227039c78631f7c0f123c7","observation_id":"4a6f70b4-328a-4871-bfd8-7e74942e7ec9","resolution":{"observed_at":"2026-08-07T10:52:49.675555Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T10:52:49.619568Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation","version":1},"reference_index":2025,"source":"pdf_text","source_observed_at":"2026-08-07T10:52:49.619568Z"},"links":{"citing_paper":"/paper/2506.04205"},"observation_digest":"sha256:63c66224b36b0c11f134ec9823946460b128126df312aaacfbaec1adadc1f4ff","observation_id":"877e747a-fb37-44e0-b48f-da98313338e3","resolution":{"observed_at":"2026-08-07T10:52:49.619568Z","resolver_source":null,"status":"parse_uncertain"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2506.04205","last_updated":"2025-06-04T17:49:10Z","latest_version":1,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-07T10:43:17.329577Z","submitted_at":"2025-06-04T17:49:10Z","title":"EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation"},"reference_resolution":{"displayed":66,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":1,"unresolved":49,"verified_exact":0,"verified_fuzzy":16},"total_outbound_references":66},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"thesis":"As of 8 August 2026, this Paper Citation Record lists 66 of 66 outbound references and 0 inbound Pith citation observations for arXiv:2506.04205."}