{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:AHTBYOKFSJBXGAQCRFP2TP5WZC","short_pith_number":"pith:AHTBYOKF","canonical_record":{"source":{"id":"2605.00161","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-30T19:31:02Z","cross_cats_sorted":[],"title_canon_sha256":"d8bed0e70c2150844e7fdc1210b7d6d552a9fd5ab9538a5e7863c65d2bea9b64","abstract_canon_sha256":"826920806f57072956b1ec4d2f9a94ce83c74969f087deea3dd8ec80c6457aee"},"schema_version":"1.0"},"canonical_sha256":"01e61c39459243730202895fa9bfb6c8adba3c2bf35d4b9c6288ede46af66422","source":{"kind":"arxiv","id":"2605.00161","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.00161","created_at":"2026-06-02T01:03:47Z"},{"alias_kind":"arxiv_version","alias_value":"2605.00161v2","created_at":"2026-06-02T01:03:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.00161","created_at":"2026-06-02T01:03:47Z"},{"alias_kind":"pith_short_12","alias_value":"AHTBYOKFSJBX","created_at":"2026-06-02T01:03:47Z"},{"alias_kind":"pith_short_16","alias_value":"AHTBYOKFSJBXGAQC","created_at":"2026-06-02T01:03:47Z"},{"alias_kind":"pith_short_8","alias_value":"AHTBYOKF","created_at":"2026-06-02T01:03:47Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:AHTBYOKFSJBXGAQCRFP2TP5WZC","target":"record","payload":{"canonical_record":{"source":{"id":"2605.00161","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-30T19:31:02Z","cross_cats_sorted":[],"title_canon_sha256":"d8bed0e70c2150844e7fdc1210b7d6d552a9fd5ab9538a5e7863c65d2bea9b64","abstract_canon_sha256":"826920806f57072956b1ec4d2f9a94ce83c74969f087deea3dd8ec80c6457aee"},"schema_version":"1.0"},"canonical_sha256":"01e61c39459243730202895fa9bfb6c8adba3c2bf35d4b9c6288ede46af66422","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:47.806932Z","signature_b64":"KZkf9mbsGtCRZqWHDxFhoCWYVJP9cmOL/4UDVUm4iVqiRunv9IFt78RHJeydy4V+ClljSBqxq1s6zuhHKLrDDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"01e61c39459243730202895fa9bfb6c8adba3c2bf35d4b9c6288ede46af66422","last_reissued_at":"2026-06-02T01:03:47.806481Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:47.806481Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.00161","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-02T01:03:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OxKz9peL7lRpv9QECv0r3A5cOB4Mnv0DLwEGxT0zM8Wx7Dg3c6bSnUXgASAdzWgQEZ+1cDef/RUXLCB1qE5HBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-28T13:39:18.638757Z"},"content_sha256":"7a5d99a94ccb788adfe2041a0b3bd363565dac51f73e143e2f1a6a6a53794bd0","schema_version":"1.0","event_id":"sha256:7a5d99a94ccb788adfe2041a0b3bd363565dac51f73e143e2f1a6a6a53794bd0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:AHTBYOKFSJBXGAQCRFP2TP5WZC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Consistent Diffusion Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A single consistency objective unifies masked and uniform discrete diffusion while delivering state-of-the-art few-step text generation.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Hasan Amin, Ming Yin, Rajiv Khanna, Subhojit Som, Xia Song, Yaser Souri, Yuan Gao","submitted_at":"2026-04-30T19:31:02Z","abstract_excerpt":"Diffusion language models (DLMs) are an attractive alternative to autoregressive models because they promise sublinear-time, parallel generation, yet practical gains remain elusive as high-quality samples still demand hundreds of refinement steps. In continuous domains, consistency training along the probability-flow ODE is a popular recipe to accelerate diffusion. For discrete diffusion, no analogous sample-space ODE exists, making direct adaptation ill-defined. We argue that the right discrete substitute is the exact posterior bridge, the closed-form conditional law linking any two noise lev"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"CDLM establishes a new state of the art on both conditional and unconditional text-generation, consistently outperforming strong base discrete diffusion models and often even multi-stage distilled baselines across sampling budgets, with the largest gains in the few-step regime.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the exact posterior bridge serves as the natural discrete substitute for the probability-flow ODE and that enforcing path-invariance in expectation across these bridges produces superior denoisers without introducing new failure modes.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"CDLM trains denoisers to be path-invariant across stochastic posterior bridges in discrete diffusion, unifying prior methods and achieving new SOTA few-step text generation performance.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A single consistency objective unifies masked and uniform discrete diffusion while delivering state-of-the-art few-step text generation.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"0975a0bd4c3569125d1239de55abb26c336afb2d0fd5e9777baeb32d5deef9d0"},"source":{"id":"2605.00161","kind":"arxiv","version":2},"verdict":{"id":"ead87992-3fde-4199-99b2-efa1249a7368","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-09T20:56:31.418642Z","strongest_claim":"CDLM establishes a new state of the art on both conditional and unconditional text-generation, consistently outperforming strong base discrete diffusion models and often even multi-stage distilled baselines across sampling budgets, with the largest gains in the few-step regime.","one_line_summary":"CDLM trains denoisers to be path-invariant across stochastic posterior bridges in discrete diffusion, unifying prior methods and achieving new SOTA few-step text generation performance.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the exact posterior bridge serves as the natural discrete substitute for the probability-flow ODE and that enforcing path-invariance in expectation across these bridges produces superior denoisers without introducing new failure modes.","pith_extraction_headline":"A single consistency objective unifies masked and uniform discrete diffusion while delivering state-of-the-art few-step text generation."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.00161/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-20T20:37:13.938567Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T18:26:16.581082Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"9392d7e32d0d861e4f5c57879526aa9bf96b47aad5b89cdec67c64a0d45e9226"},"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":"ead87992-3fde-4199-99b2-efa1249a7368"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-02T01:03:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"e0iddTWGRjahX1+uS4mpjDtzdzrPd498n0U+3zA9pjknthMusjE3+XuvcEbNNKfhNeso+BgbFsVR+jYuqB+qBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-28T13:39:18.639235Z"},"content_sha256":"38fe4e35dc692196492c745ffb624792302db27cca0e267f961d99f496107ff0","schema_version":"1.0","event_id":"sha256:38fe4e35dc692196492c745ffb624792302db27cca0e267f961d99f496107ff0"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/AHTBYOKFSJBXGAQCRFP2TP5WZC/bundle.json","state_url":"https://pith.science/pith/AHTBYOKFSJBXGAQCRFP2TP5WZC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/AHTBYOKFSJBXGAQCRFP2TP5WZC/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":"2026-06-28T13:39:18Z","links":{"resolver":"https://pith.science/pith/AHTBYOKFSJBXGAQCRFP2TP5WZC","bundle":"https://pith.science/pith/AHTBYOKFSJBXGAQCRFP2TP5WZC/bundle.json","state":"https://pith.science/pith/AHTBYOKFSJBXGAQCRFP2TP5WZC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/AHTBYOKFSJBXGAQCRFP2TP5WZC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:AHTBYOKFSJBXGAQCRFP2TP5WZC","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":"826920806f57072956b1ec4d2f9a94ce83c74969f087deea3dd8ec80c6457aee","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-30T19:31:02Z","title_canon_sha256":"d8bed0e70c2150844e7fdc1210b7d6d552a9fd5ab9538a5e7863c65d2bea9b64"},"schema_version":"1.0","source":{"id":"2605.00161","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.00161","created_at":"2026-06-02T01:03:47Z"},{"alias_kind":"arxiv_version","alias_value":"2605.00161v2","created_at":"2026-06-02T01:03:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.00161","created_at":"2026-06-02T01:03:47Z"},{"alias_kind":"pith_short_12","alias_value":"AHTBYOKFSJBX","created_at":"2026-06-02T01:03:47Z"},{"alias_kind":"pith_short_16","alias_value":"AHTBYOKFSJBXGAQC","created_at":"2026-06-02T01:03:47Z"},{"alias_kind":"pith_short_8","alias_value":"AHTBYOKF","created_at":"2026-06-02T01:03:47Z"}],"graph_snapshots":[{"event_id":"sha256:38fe4e35dc692196492c745ffb624792302db27cca0e267f961d99f496107ff0","target":"graph","created_at":"2026-06-02T01:03:47Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"CDLM establishes a new state of the art on both conditional and unconditional text-generation, consistently outperforming strong base discrete diffusion models and often even multi-stage distilled baselines across sampling budgets, with the largest gains in the few-step regime."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the exact posterior bridge serves as the natural discrete substitute for the probability-flow ODE and that enforcing path-invariance in expectation across these bridges produces superior denoisers without introducing new failure modes."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"CDLM trains denoisers to be path-invariant across stochastic posterior bridges in discrete diffusion, unifying prior methods and achieving new SOTA few-step text generation performance."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A single consistency objective unifies masked and uniform discrete diffusion while delivering state-of-the-art few-step text generation."}],"snapshot_sha256":"0975a0bd4c3569125d1239de55abb26c336afb2d0fd5e9777baeb32d5deef9d0"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-20T20:37:13.938567Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T18:26:16.581082Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2605.00161/integrity.json","findings":[],"snapshot_sha256":"9392d7e32d0d861e4f5c57879526aa9bf96b47aad5b89cdec67c64a0d45e9226","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Diffusion language models (DLMs) are an attractive alternative to autoregressive models because they promise sublinear-time, parallel generation, yet practical gains remain elusive as high-quality samples still demand hundreds of refinement steps. In continuous domains, consistency training along the probability-flow ODE is a popular recipe to accelerate diffusion. For discrete diffusion, no analogous sample-space ODE exists, making direct adaptation ill-defined. We argue that the right discrete substitute is the exact posterior bridge, the closed-form conditional law linking any two noise lev","authors_text":"Hasan Amin, Ming Yin, Rajiv Khanna, Subhojit Som, Xia Song, Yaser Souri, Yuan Gao","cross_cats":[],"headline":"A single consistency objective unifies masked and uniform discrete diffusion while delivering state-of-the-art few-step text generation.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-30T19:31:02Z","title":"Consistent Diffusion Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.00161","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-09T20:56:31.418642Z","id":"ead87992-3fde-4199-99b2-efa1249a7368","model_set":{"reader":"grok-4.3"},"one_line_summary":"CDLM trains denoisers to be path-invariant across stochastic posterior bridges in discrete diffusion, unifying prior methods and achieving new SOTA few-step text generation performance.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A single consistency objective unifies masked and uniform discrete diffusion while delivering state-of-the-art few-step text generation.","strongest_claim":"CDLM establishes a new state of the art on both conditional and unconditional text-generation, consistently outperforming strong base discrete diffusion models and often even multi-stage distilled baselines across sampling budgets, with the largest gains in the few-step regime.","weakest_assumption":"That the exact posterior bridge serves as the natural discrete substitute for the probability-flow ODE and that enforcing path-invariance in expectation across these bridges produces superior denoisers without introducing new failure modes."}},"verdict_id":"ead87992-3fde-4199-99b2-efa1249a7368"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:7a5d99a94ccb788adfe2041a0b3bd363565dac51f73e143e2f1a6a6a53794bd0","target":"record","created_at":"2026-06-02T01:03:47Z","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":"826920806f57072956b1ec4d2f9a94ce83c74969f087deea3dd8ec80c6457aee","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-30T19:31:02Z","title_canon_sha256":"d8bed0e70c2150844e7fdc1210b7d6d552a9fd5ab9538a5e7863c65d2bea9b64"},"schema_version":"1.0","source":{"id":"2605.00161","kind":"arxiv","version":2}},"canonical_sha256":"01e61c39459243730202895fa9bfb6c8adba3c2bf35d4b9c6288ede46af66422","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"01e61c39459243730202895fa9bfb6c8adba3c2bf35d4b9c6288ede46af66422","first_computed_at":"2026-06-02T01:03:47.806481Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T01:03:47.806481Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"KZkf9mbsGtCRZqWHDxFhoCWYVJP9cmOL/4UDVUm4iVqiRunv9IFt78RHJeydy4V+ClljSBqxq1s6zuhHKLrDDA==","signature_status":"signed_v1","signed_at":"2026-06-02T01:03:47.806932Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.00161","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7a5d99a94ccb788adfe2041a0b3bd363565dac51f73e143e2f1a6a6a53794bd0","sha256:38fe4e35dc692196492c745ffb624792302db27cca0e267f961d99f496107ff0"],"state_sha256":"a645639237ad9c8b011de06213e69a93795bdd4d9fa88583f4ddbeefd9ee8cbe"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oRLVmiA1Nvl8GRjgydEqF5mmxXlsaZ8vrP+qqQS20l4MMu8Ac/afZkz6RAndFpnIQxCbzQluRGS+f4lGpoq3DQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-28T13:39:18.641993Z","bundle_sha256":"fb3c0f6d5eceac418c7f2f2ea780bf3cf76a2fb5e7db7f54176d00869f8c31da"}}