{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:5Y77NGQWSQ7T7MVZKT5ZAAIIEG","short_pith_number":"pith:5Y77NGQW","canonical_record":{"source":{"id":"2404.11500","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-17T15:53:49Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"94d5db275811cd2c4b1b786b3a771397fcc5c6091fbbcccfda634c876512a19a","abstract_canon_sha256":"2a22d0884a9ac1843da92e38e23c6af883aa7541b7953654c178ba507149e4b5"},"schema_version":"1.0"},"canonical_sha256":"ee3ff69a16943f3fb2b954fb90010821aeaed31e152dd41f853e82eece127103","source":{"kind":"arxiv","id":"2404.11500","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.11500","created_at":"2026-07-05T08:09:09Z"},{"alias_kind":"arxiv_version","alias_value":"2404.11500v1","created_at":"2026-07-05T08:09:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.11500","created_at":"2026-07-05T08:09:09Z"},{"alias_kind":"pith_short_12","alias_value":"5Y77NGQWSQ7T","created_at":"2026-07-05T08:09:09Z"},{"alias_kind":"pith_short_16","alias_value":"5Y77NGQWSQ7T7MVZ","created_at":"2026-07-05T08:09:09Z"},{"alias_kind":"pith_short_8","alias_value":"5Y77NGQW","created_at":"2026-07-05T08:09:09Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:5Y77NGQWSQ7T7MVZKT5ZAAIIEG","target":"record","payload":{"canonical_record":{"source":{"id":"2404.11500","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-17T15:53:49Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"94d5db275811cd2c4b1b786b3a771397fcc5c6091fbbcccfda634c876512a19a","abstract_canon_sha256":"2a22d0884a9ac1843da92e38e23c6af883aa7541b7953654c178ba507149e4b5"},"schema_version":"1.0"},"canonical_sha256":"ee3ff69a16943f3fb2b954fb90010821aeaed31e152dd41f853e82eece127103","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:09:09.611116Z","signature_b64":"vAvTssa5QS6lE2xOb7FR1JGuTkHKRpw34kKnMi+DccnKUJkmRNCfG9S5Va+3Ku10Idfm2u7UgQZxpdanLX1+DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ee3ff69a16943f3fb2b954fb90010821aeaed31e152dd41f853e82eece127103","last_reissued_at":"2026-07-05T08:09:09.610674Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:09:09.610674Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2404.11500","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:09:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4VQ8I0WY25kQ9LxeowUsY77H3WcgfuJARIT7d/2AOJtFDuZfcnraH1SRat9vISXZwTySGny9eqUFPtpx2Ca0Bw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T07:11:56.443076Z"},"content_sha256":"09cc7f83624fbe388c7b683a51ed1ae7e07854c05f26c82ee613b9748ec70c02","schema_version":"1.0","event_id":"sha256:09cc7f83624fbe388c7b683a51ed1ae7e07854c05f26c82ee613b9748ec70c02"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:5Y77NGQWSQ7T7MVZKT5ZAAIIEG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Paraphrase and Solve: Exploring and Exploiting the Impact of Surface Form on Mathematical Reasoning in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Diego Antognini, Yada Zhu, Yang Zhang, Yoon Kim, Yue Zhou","submitted_at":"2024-04-17T15:53:49Z","abstract_excerpt":"This paper studies the relationship between the surface form of a mathematical problem and its solvability by large language models. We find that subtle alterations in the surface form can significantly impact the answer distribution and the solve rate, exposing the language model's lack of robustness and sensitivity to the surface form in reasoning through complex problems. To improve mathematical reasoning performance, we propose Self-Consistency-over-Paraphrases (SCoP), which diversifies reasoning paths from specific surface forms of the problem. We evaluate our approach on four mathematics"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.11500","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2404.11500/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:09:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wzEMmC0ZyckM+Lf2Sd4apJlfZXG25IfLi/zI3WUeNR/gxdRsBF6eu/92N6J1KbW9fvGczUAU+xR5IArCPe/lBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T07:11:56.443799Z"},"content_sha256":"cec0a0952bc73ed53ecedb80309190c09614c20690af773733a42c21c856bd91","schema_version":"1.0","event_id":"sha256:cec0a0952bc73ed53ecedb80309190c09614c20690af773733a42c21c856bd91"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/5Y77NGQWSQ7T7MVZKT5ZAAIIEG/bundle.json","state_url":"https://pith.science/pith/5Y77NGQWSQ7T7MVZKT5ZAAIIEG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/5Y77NGQWSQ7T7MVZKT5ZAAIIEG/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-09T07:11:56Z","links":{"resolver":"https://pith.science/pith/5Y77NGQWSQ7T7MVZKT5ZAAIIEG","bundle":"https://pith.science/pith/5Y77NGQWSQ7T7MVZKT5ZAAIIEG/bundle.json","state":"https://pith.science/pith/5Y77NGQWSQ7T7MVZKT5ZAAIIEG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/5Y77NGQWSQ7T7MVZKT5ZAAIIEG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:5Y77NGQWSQ7T7MVZKT5ZAAIIEG","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"2a22d0884a9ac1843da92e38e23c6af883aa7541b7953654c178ba507149e4b5","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-17T15:53:49Z","title_canon_sha256":"94d5db275811cd2c4b1b786b3a771397fcc5c6091fbbcccfda634c876512a19a"},"schema_version":"1.0","source":{"id":"2404.11500","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.11500","created_at":"2026-07-05T08:09:09Z"},{"alias_kind":"arxiv_version","alias_value":"2404.11500v1","created_at":"2026-07-05T08:09:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.11500","created_at":"2026-07-05T08:09:09Z"},{"alias_kind":"pith_short_12","alias_value":"5Y77NGQWSQ7T","created_at":"2026-07-05T08:09:09Z"},{"alias_kind":"pith_short_16","alias_value":"5Y77NGQWSQ7T7MVZ","created_at":"2026-07-05T08:09:09Z"},{"alias_kind":"pith_short_8","alias_value":"5Y77NGQW","created_at":"2026-07-05T08:09:09Z"}],"graph_snapshots":[{"event_id":"sha256:cec0a0952bc73ed53ecedb80309190c09614c20690af773733a42c21c856bd91","target":"graph","created_at":"2026-07-05T08:09:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2404.11500/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper studies the relationship between the surface form of a mathematical problem and its solvability by large language models. We find that subtle alterations in the surface form can significantly impact the answer distribution and the solve rate, exposing the language model's lack of robustness and sensitivity to the surface form in reasoning through complex problems. To improve mathematical reasoning performance, we propose Self-Consistency-over-Paraphrases (SCoP), which diversifies reasoning paths from specific surface forms of the problem. We evaluate our approach on four mathematics","authors_text":"Diego Antognini, Yada Zhu, Yang Zhang, Yoon Kim, Yue Zhou","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-17T15:53:49Z","title":"Paraphrase and Solve: Exploring and Exploiting the Impact of Surface Form on Mathematical Reasoning in Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.11500","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:09cc7f83624fbe388c7b683a51ed1ae7e07854c05f26c82ee613b9748ec70c02","target":"record","created_at":"2026-07-05T08:09:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"2a22d0884a9ac1843da92e38e23c6af883aa7541b7953654c178ba507149e4b5","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-17T15:53:49Z","title_canon_sha256":"94d5db275811cd2c4b1b786b3a771397fcc5c6091fbbcccfda634c876512a19a"},"schema_version":"1.0","source":{"id":"2404.11500","kind":"arxiv","version":1}},"canonical_sha256":"ee3ff69a16943f3fb2b954fb90010821aeaed31e152dd41f853e82eece127103","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ee3ff69a16943f3fb2b954fb90010821aeaed31e152dd41f853e82eece127103","first_computed_at":"2026-07-05T08:09:09.610674Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:09:09.610674Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vAvTssa5QS6lE2xOb7FR1JGuTkHKRpw34kKnMi+DccnKUJkmRNCfG9S5Va+3Ku10Idfm2u7UgQZxpdanLX1+DQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:09:09.611116Z","signed_message":"canonical_sha256_bytes"},"source_id":"2404.11500","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:09cc7f83624fbe388c7b683a51ed1ae7e07854c05f26c82ee613b9748ec70c02","sha256:cec0a0952bc73ed53ecedb80309190c09614c20690af773733a42c21c856bd91"],"state_sha256":"0291ec35e7117a7e13699c5516e8319085551868c6086c459fd787803f6de616"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xL+/r3pspMrqjNvNe3cizsi/wqIYT310opmuCTMgN2nq3HszOq6IWcXxEPgfQ59v3w2cqQLP1Cgdc6oX9DGYBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T07:11:56.448044Z","bundle_sha256":"54be3dddd2496e3435a1652f1d0c514ecc7cc0ab1cde294a5e32400781e9a98d"}}