{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:RJZFVMRBQEMFGGUMYCM5GN4WV2","short_pith_number":"pith:RJZFVMRB","canonical_record":{"source":{"id":"2505.23829","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T08:09:10Z","cross_cats_sorted":[],"title_canon_sha256":"f56257fecb94a6155d1cc23c7b4f3ae011945be4290279ae02de084dc4bf9340","abstract_canon_sha256":"2fda1b317296174ad39954dbecb21e5cfec11e503ba179bef64ed4e2bb7ab5fc"},"schema_version":"1.0"},"canonical_sha256":"8a725ab2218118531a8cc099d33796ae9917536e28d44c66ef4bc09f556e493f","source":{"kind":"arxiv","id":"2505.23829","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.23829","created_at":"2026-07-05T11:12:36Z"},{"alias_kind":"arxiv_version","alias_value":"2505.23829v1","created_at":"2026-07-05T11:12:36Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.23829","created_at":"2026-07-05T11:12:36Z"},{"alias_kind":"pith_short_12","alias_value":"RJZFVMRBQEMF","created_at":"2026-07-05T11:12:36Z"},{"alias_kind":"pith_short_16","alias_value":"RJZFVMRBQEMFGGUM","created_at":"2026-07-05T11:12:36Z"},{"alias_kind":"pith_short_8","alias_value":"RJZFVMRB","created_at":"2026-07-05T11:12:36Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:RJZFVMRBQEMFGGUMYCM5GN4WV2","target":"record","payload":{"canonical_record":{"source":{"id":"2505.23829","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T08:09:10Z","cross_cats_sorted":[],"title_canon_sha256":"f56257fecb94a6155d1cc23c7b4f3ae011945be4290279ae02de084dc4bf9340","abstract_canon_sha256":"2fda1b317296174ad39954dbecb21e5cfec11e503ba179bef64ed4e2bb7ab5fc"},"schema_version":"1.0"},"canonical_sha256":"8a725ab2218118531a8cc099d33796ae9917536e28d44c66ef4bc09f556e493f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:12:36.095824Z","signature_b64":"6MWPQ1NAxDDyoYayR5+pdsGNSB+UIpw/5stisRtB2Ezv/DC9+ey0NhpQfP6kbbmKh8QSx2MNyjFhmM9tzDLZAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8a725ab2218118531a8cc099d33796ae9917536e28d44c66ef4bc09f556e493f","last_reissued_at":"2026-07-05T11:12:36.095272Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:12:36.095272Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.23829","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":"2026-07-05T11:12:36Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wVDXeNL5B5yoE+6qmxccRSIrJ9+/bv1marbe4BBlr57TwabwMIlUVA6zV12MWRcTIok5vJd7RxQh+fIKNEG/CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T12:50:19.117027Z"},"content_sha256":"894b2822ee08e155644601e49053248a33112ac88735d904fdecf7a9b8658880","schema_version":"1.0","event_id":"sha256:894b2822ee08e155644601e49053248a33112ac88735d904fdecf7a9b8658880"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:RJZFVMRBQEMFGGUMYCM5GN4WV2","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"BiasFilter: An Inference-Time Debiasing Framework for Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hongying Zan, Min Peng, Ruizhe Chen, Xiaoqing Cheng, Yuxiang Jia","submitted_at":"2025-05-28T08:09:10Z","abstract_excerpt":"Mitigating social bias in large language models (LLMs) has become an increasingly important research objective. However, existing debiasing methods often incur high human and computational costs, exhibit limited effectiveness, and struggle to scale to larger models and open-ended generation tasks. To address these limitations, this paper proposes BiasFilter, a model-agnostic, inference-time debiasing framework that integrates seamlessly with both open-source and API-based LLMs. Instead of relying on retraining with balanced data or modifying model parameters, BiasFilter enforces fairness by fi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.23829","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/2505.23829/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":"2026-07-05T11:12:36Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ojJQhDID4KJyMcn9WdsLZAw3xrAzsw0p3UNoEgg1SHWIk1luM55zb+RFVaO9YbzDBwdGLWUdof/C6gE88s7GBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T12:50:19.117935Z"},"content_sha256":"cfb1f0f5a28d154b7843a92df3fc4e4d6df87a6f4c4dd7781a8979bf3296e847","schema_version":"1.0","event_id":"sha256:cfb1f0f5a28d154b7843a92df3fc4e4d6df87a6f4c4dd7781a8979bf3296e847"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RJZFVMRBQEMFGGUMYCM5GN4WV2/bundle.json","state_url":"https://pith.science/pith/RJZFVMRBQEMFGGUMYCM5GN4WV2/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RJZFVMRBQEMFGGUMYCM5GN4WV2/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":"2026-08-09T12:50:19Z","links":{"resolver":"https://pith.science/pith/RJZFVMRBQEMFGGUMYCM5GN4WV2","bundle":"https://pith.science/pith/RJZFVMRBQEMFGGUMYCM5GN4WV2/bundle.json","state":"https://pith.science/pith/RJZFVMRBQEMFGGUMYCM5GN4WV2/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RJZFVMRBQEMFGGUMYCM5GN4WV2/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:RJZFVMRBQEMFGGUMYCM5GN4WV2","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":"2fda1b317296174ad39954dbecb21e5cfec11e503ba179bef64ed4e2bb7ab5fc","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T08:09:10Z","title_canon_sha256":"f56257fecb94a6155d1cc23c7b4f3ae011945be4290279ae02de084dc4bf9340"},"schema_version":"1.0","source":{"id":"2505.23829","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.23829","created_at":"2026-07-05T11:12:36Z"},{"alias_kind":"arxiv_version","alias_value":"2505.23829v1","created_at":"2026-07-05T11:12:36Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.23829","created_at":"2026-07-05T11:12:36Z"},{"alias_kind":"pith_short_12","alias_value":"RJZFVMRBQEMF","created_at":"2026-07-05T11:12:36Z"},{"alias_kind":"pith_short_16","alias_value":"RJZFVMRBQEMFGGUM","created_at":"2026-07-05T11:12:36Z"},{"alias_kind":"pith_short_8","alias_value":"RJZFVMRB","created_at":"2026-07-05T11:12:36Z"}],"graph_snapshots":[{"event_id":"sha256:cfb1f0f5a28d154b7843a92df3fc4e4d6df87a6f4c4dd7781a8979bf3296e847","target":"graph","created_at":"2026-07-05T11:12:36Z","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/2505.23829/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Mitigating social bias in large language models (LLMs) has become an increasingly important research objective. However, existing debiasing methods often incur high human and computational costs, exhibit limited effectiveness, and struggle to scale to larger models and open-ended generation tasks. To address these limitations, this paper proposes BiasFilter, a model-agnostic, inference-time debiasing framework that integrates seamlessly with both open-source and API-based LLMs. Instead of relying on retraining with balanced data or modifying model parameters, BiasFilter enforces fairness by fi","authors_text":"Hongying Zan, Min Peng, Ruizhe Chen, Xiaoqing Cheng, Yuxiang Jia","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T08:09:10Z","title":"BiasFilter: An Inference-Time Debiasing Framework for Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.23829","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":"sha256:894b2822ee08e155644601e49053248a33112ac88735d904fdecf7a9b8658880","target":"record","created_at":"2026-07-05T11:12:36Z","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":"2fda1b317296174ad39954dbecb21e5cfec11e503ba179bef64ed4e2bb7ab5fc","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-28T08:09:10Z","title_canon_sha256":"f56257fecb94a6155d1cc23c7b4f3ae011945be4290279ae02de084dc4bf9340"},"schema_version":"1.0","source":{"id":"2505.23829","kind":"arxiv","version":1}},"canonical_sha256":"8a725ab2218118531a8cc099d33796ae9917536e28d44c66ef4bc09f556e493f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8a725ab2218118531a8cc099d33796ae9917536e28d44c66ef4bc09f556e493f","first_computed_at":"2026-07-05T11:12:36.095272Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:12:36.095272Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"6MWPQ1NAxDDyoYayR5+pdsGNSB+UIpw/5stisRtB2Ezv/DC9+ey0NhpQfP6kbbmKh8QSx2MNyjFhmM9tzDLZAw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:12:36.095824Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.23829","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:894b2822ee08e155644601e49053248a33112ac88735d904fdecf7a9b8658880","sha256:cfb1f0f5a28d154b7843a92df3fc4e4d6df87a6f4c4dd7781a8979bf3296e847"],"state_sha256":"f2d8eaf0fd446bee08db1c68fd14f8e0ce97176e0ac7e8554b8c4b63bfce0b0d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uTOmrLJ4o+Vzy35yvSTYID/Fy7tL0tdJ9akOjTnF2HcGze63yncToLNk/y2EUYIr2nuyaKTuXYPSm4hL3/neAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T12:50:19.124639Z","bundle_sha256":"14aa6efd5d319365b5e614234125c0b5be1f0c234066a166739f93b99da19af0"}}