{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:26OFLDTLSEV7LRK74ZLXOKQBGI","short_pith_number":"pith:26OFLDTL","canonical_record":{"source":{"id":"2605.29398","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-28T05:47:40Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"dde5a7c57de24e903572613750bf63f443ac80ee37e380f7b6474f992581ad8e","abstract_canon_sha256":"ffbf0ffeaefd483e7f45bb3473c1f7e1473c27f7419bfc15da877e91324039de"},"schema_version":"1.0"},"canonical_sha256":"d79c558e6b912bf5c55fe657772a01321ca17288cd8a07ece8817c6b5e2850a7","source":{"kind":"arxiv","id":"2605.29398","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.29398","created_at":"2026-05-29T01:05:37Z"},{"alias_kind":"arxiv_version","alias_value":"2605.29398v1","created_at":"2026-05-29T01:05:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.29398","created_at":"2026-05-29T01:05:37Z"},{"alias_kind":"pith_short_12","alias_value":"26OFLDTLSEV7","created_at":"2026-05-29T01:05:37Z"},{"alias_kind":"pith_short_16","alias_value":"26OFLDTLSEV7LRK7","created_at":"2026-05-29T01:05:37Z"},{"alias_kind":"pith_short_8","alias_value":"26OFLDTL","created_at":"2026-05-29T01:05:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:26OFLDTLSEV7LRK74ZLXOKQBGI","target":"record","payload":{"canonical_record":{"source":{"id":"2605.29398","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-28T05:47:40Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"dde5a7c57de24e903572613750bf63f443ac80ee37e380f7b6474f992581ad8e","abstract_canon_sha256":"ffbf0ffeaefd483e7f45bb3473c1f7e1473c27f7419bfc15da877e91324039de"},"schema_version":"1.0"},"canonical_sha256":"d79c558e6b912bf5c55fe657772a01321ca17288cd8a07ece8817c6b5e2850a7","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T01:05:37.022187Z","signature_b64":"L9kbkS6nZYpXHVLu5f6trxmWHYxt06PKvGQX5O0G9az8N2BaXwRWAifjtutv79CtbK/NPyZ6mCQrNi6RgNiOAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d79c558e6b912bf5c55fe657772a01321ca17288cd8a07ece8817c6b5e2850a7","last_reissued_at":"2026-05-29T01:05:37.021404Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T01:05:37.021404Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.29398","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":"2026-05-29T01:05:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"53u6f5pZibGIZgFFSbxCCCF7C4oW6XfGXLV1BRXqSLvKMDv/1avSrlXESSUdWjInZmfxu+/amBytDaYTfUelBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T20:29:53.953106Z"},"content_sha256":"4ef5f66b7b349ce3a94385b135b31a2ff49cde70008f16e193a4c090f4fd39ed","schema_version":"1.0","event_id":"sha256:4ef5f66b7b349ce3a94385b135b31a2ff49cde70008f16e193a4c090f4fd39ed"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:26OFLDTLSEV7LRK74ZLXOKQBGI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Che Liu, Ilija Bogunovic, Keyue Jiang, Qifang Zhao, Sangwoong Yoon, Xiaohang Tang, Xiaoxiao Xu","submitted_at":"2026-05-28T05:47:40Z","abstract_excerpt":"Reinforcement learning (RL) can be used to improve the policy (denoiser) of diffusion large language models (dLLMs), while being hindered by the intractability of the policy likelihood. A dominant and efficient family of methods replaces the likelihood in standard RL with its evidence lower bound (ELBO), estimated from randomly masked sequences. Despite being well aligned with pre-training, these approaches introduce bias through training--inference mismatch by using the ELBO as a likelihood surrogate, which can degrade performance. In this work, we propose Guided Denoiser Self-Distillation (G"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.29398","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/2605.29398/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":"2026-05-29T01:05:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"saa7VWWbKV/DuNS7+u28Uue2JD3mwwBM53B+vMAVRUAZkrjdRDcCXUa4BW0ZagKzM7UFfgwQQFXuofGO1BzQAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T20:29:53.953519Z"},"content_sha256":"9ed9fb2c64e636efec74c10781b075ee64b65fd25e83fb98d347c93dd12b41da","schema_version":"1.0","event_id":"sha256:9ed9fb2c64e636efec74c10781b075ee64b65fd25e83fb98d347c93dd12b41da"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/26OFLDTLSEV7LRK74ZLXOKQBGI/bundle.json","state_url":"https://pith.science/pith/26OFLDTLSEV7LRK74ZLXOKQBGI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/26OFLDTLSEV7LRK74ZLXOKQBGI/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":"2026-06-02T20:29:53Z","links":{"resolver":"https://pith.science/pith/26OFLDTLSEV7LRK74ZLXOKQBGI","bundle":"https://pith.science/pith/26OFLDTLSEV7LRK74ZLXOKQBGI/bundle.json","state":"https://pith.science/pith/26OFLDTLSEV7LRK74ZLXOKQBGI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/26OFLDTLSEV7LRK74ZLXOKQBGI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:26OFLDTLSEV7LRK74ZLXOKQBGI","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":"ffbf0ffeaefd483e7f45bb3473c1f7e1473c27f7419bfc15da877e91324039de","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-28T05:47:40Z","title_canon_sha256":"dde5a7c57de24e903572613750bf63f443ac80ee37e380f7b6474f992581ad8e"},"schema_version":"1.0","source":{"id":"2605.29398","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.29398","created_at":"2026-05-29T01:05:37Z"},{"alias_kind":"arxiv_version","alias_value":"2605.29398v1","created_at":"2026-05-29T01:05:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.29398","created_at":"2026-05-29T01:05:37Z"},{"alias_kind":"pith_short_12","alias_value":"26OFLDTLSEV7","created_at":"2026-05-29T01:05:37Z"},{"alias_kind":"pith_short_16","alias_value":"26OFLDTLSEV7LRK7","created_at":"2026-05-29T01:05:37Z"},{"alias_kind":"pith_short_8","alias_value":"26OFLDTL","created_at":"2026-05-29T01:05:37Z"}],"graph_snapshots":[{"event_id":"sha256:9ed9fb2c64e636efec74c10781b075ee64b65fd25e83fb98d347c93dd12b41da","target":"graph","created_at":"2026-05-29T01:05:37Z","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/2605.29398/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement learning (RL) can be used to improve the policy (denoiser) of diffusion large language models (dLLMs), while being hindered by the intractability of the policy likelihood. A dominant and efficient family of methods replaces the likelihood in standard RL with its evidence lower bound (ELBO), estimated from randomly masked sequences. Despite being well aligned with pre-training, these approaches introduce bias through training--inference mismatch by using the ELBO as a likelihood surrogate, which can degrade performance. In this work, we propose Guided Denoiser Self-Distillation (G","authors_text":"Che Liu, Ilija Bogunovic, Keyue Jiang, Qifang Zhao, Sangwoong Yoon, Xiaohang Tang, Xiaoxiao Xu","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-28T05:47:40Z","title":"GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.29398","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":"sha256:4ef5f66b7b349ce3a94385b135b31a2ff49cde70008f16e193a4c090f4fd39ed","target":"record","created_at":"2026-05-29T01:05:37Z","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":"ffbf0ffeaefd483e7f45bb3473c1f7e1473c27f7419bfc15da877e91324039de","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-28T05:47:40Z","title_canon_sha256":"dde5a7c57de24e903572613750bf63f443ac80ee37e380f7b6474f992581ad8e"},"schema_version":"1.0","source":{"id":"2605.29398","kind":"arxiv","version":1}},"canonical_sha256":"d79c558e6b912bf5c55fe657772a01321ca17288cd8a07ece8817c6b5e2850a7","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d79c558e6b912bf5c55fe657772a01321ca17288cd8a07ece8817c6b5e2850a7","first_computed_at":"2026-05-29T01:05:37.021404Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-29T01:05:37.021404Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"L9kbkS6nZYpXHVLu5f6trxmWHYxt06PKvGQX5O0G9az8N2BaXwRWAifjtutv79CtbK/NPyZ6mCQrNi6RgNiOAw==","signature_status":"signed_v1","signed_at":"2026-05-29T01:05:37.022187Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.29398","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4ef5f66b7b349ce3a94385b135b31a2ff49cde70008f16e193a4c090f4fd39ed","sha256:9ed9fb2c64e636efec74c10781b075ee64b65fd25e83fb98d347c93dd12b41da"],"state_sha256":"09043678d08119a0f708683cfcf24b4c791ca22ce94d783fd1d3a98daed7025c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZkBo+5zfMEDNS04YBa7DkA7tD4MsQDyCMOlbUlcl4gDdD1WeKJdKWYVBsXz6+rw0ol4XCq06YNtUkkM2ND+ODg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-02T20:29:53.955941Z","bundle_sha256":"0f45efda71d3597e60a4440919b19b2256d21f381b1d334120f402d52110c77f"}}