{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:VBPYWGYHBY5HRW2SAKCDVO3CQO","short_pith_number":"pith:VBPYWGYH","canonical_record":{"source":{"id":"2412.07675","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-10T17:02:58Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"14bf03d5a9347ab02c6a56d6fdcbd5d7e63e4d6daac3b8eb2e599a03c15f93b8","abstract_canon_sha256":"53fdad395a9eb6347cb1b03353d1ebe6e69a0ba407e323acf4fda68854661116"},"schema_version":"1.0"},"canonical_sha256":"a85f8b1b070e3a78db5202843abb628380e3a9a4325b787c07c6062936dae628","source":{"kind":"arxiv","id":"2412.07675","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.07675","created_at":"2026-07-05T09:51:39Z"},{"alias_kind":"arxiv_version","alias_value":"2412.07675v3","created_at":"2026-07-05T09:51:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.07675","created_at":"2026-07-05T09:51:39Z"},{"alias_kind":"pith_short_12","alias_value":"VBPYWGYHBY5H","created_at":"2026-07-05T09:51:39Z"},{"alias_kind":"pith_short_16","alias_value":"VBPYWGYHBY5HRW2S","created_at":"2026-07-05T09:51:39Z"},{"alias_kind":"pith_short_8","alias_value":"VBPYWGYH","created_at":"2026-07-05T09:51:39Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:VBPYWGYHBY5HRW2SAKCDVO3CQO","target":"record","payload":{"canonical_record":{"source":{"id":"2412.07675","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-10T17:02:58Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"14bf03d5a9347ab02c6a56d6fdcbd5d7e63e4d6daac3b8eb2e599a03c15f93b8","abstract_canon_sha256":"53fdad395a9eb6347cb1b03353d1ebe6e69a0ba407e323acf4fda68854661116"},"schema_version":"1.0"},"canonical_sha256":"a85f8b1b070e3a78db5202843abb628380e3a9a4325b787c07c6062936dae628","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:51:39.363020Z","signature_b64":"elVp51yYLmcwY6zs5LBsfRiepQtqVx5As+C4L3nieaOlMX3tR/S1HnqXqBRS6AL9KC2TFIIeoCBCvGr8He0rDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a85f8b1b070e3a78db5202843abb628380e3a9a4325b787c07c6062936dae628","last_reissued_at":"2026-07-05T09:51:39.362589Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:51:39.362589Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.07675","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:51:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"C8y6yIYAfLY8iQKV1sGi9ZC7ksn02+pqQSVeks1jruo/KctXKkWWaSt+FBKyd92NcSocJpbw+gwi8hLHCRjACA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T18:16:19.772492Z"},"content_sha256":"efb58da0370b46c2ed029eae589d51c9275f929f03db27741375a75da59aabe3","schema_version":"1.0","event_id":"sha256:efb58da0370b46c2ed029eae589d51c9275f929f03db27741375a75da59aabe3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:VBPYWGYHBY5HRW2SAKCDVO3CQO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"RAZOR: Sharpening Knowledge by Cutting Bias with Unsupervised Text Rewriting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Bardh Prenkaj, Gjergji Kasneci, Shuo Yang","submitted_at":"2024-12-10T17:02:58Z","abstract_excerpt":"Despite the widespread use of LLMs due to their superior performance in various tasks, their high computational costs often lead potential users to opt for the pretraining-finetuning pipeline. However, biases prevalent in manually constructed datasets can introduce spurious correlations between tokens and labels, creating so-called shortcuts and hindering the generalizability of fine-tuned models. Existing debiasing methods often rely on prior knowledge of specific dataset biases, which is challenging to acquire a priori. We propose RAZOR (Rewriting And Zero-bias Optimization Refinement), a no"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.07675","kind":"arxiv","version":3},"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/2412.07675/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":"2026-07-05T09:51:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6cTDnWDGMNRwbsTte14BTspwA/FG9791TPAqCtPII+Ym7TLFK80fOsLehnH8v2/zpqA1mvxsUmPfqsS8z5yVBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T18:16:19.772987Z"},"content_sha256":"fecb90bbaef830f844cd34412e850a03adf928676974692481af0d9953600d46","schema_version":"1.0","event_id":"sha256:fecb90bbaef830f844cd34412e850a03adf928676974692481af0d9953600d46"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VBPYWGYHBY5HRW2SAKCDVO3CQO/bundle.json","state_url":"https://pith.science/pith/VBPYWGYHBY5HRW2SAKCDVO3CQO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VBPYWGYHBY5HRW2SAKCDVO3CQO/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":"2026-08-10T18:16:19Z","links":{"resolver":"https://pith.science/pith/VBPYWGYHBY5HRW2SAKCDVO3CQO","bundle":"https://pith.science/pith/VBPYWGYHBY5HRW2SAKCDVO3CQO/bundle.json","state":"https://pith.science/pith/VBPYWGYHBY5HRW2SAKCDVO3CQO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VBPYWGYHBY5HRW2SAKCDVO3CQO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:VBPYWGYHBY5HRW2SAKCDVO3CQO","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":"53fdad395a9eb6347cb1b03353d1ebe6e69a0ba407e323acf4fda68854661116","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-10T17:02:58Z","title_canon_sha256":"14bf03d5a9347ab02c6a56d6fdcbd5d7e63e4d6daac3b8eb2e599a03c15f93b8"},"schema_version":"1.0","source":{"id":"2412.07675","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.07675","created_at":"2026-07-05T09:51:39Z"},{"alias_kind":"arxiv_version","alias_value":"2412.07675v3","created_at":"2026-07-05T09:51:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.07675","created_at":"2026-07-05T09:51:39Z"},{"alias_kind":"pith_short_12","alias_value":"VBPYWGYHBY5H","created_at":"2026-07-05T09:51:39Z"},{"alias_kind":"pith_short_16","alias_value":"VBPYWGYHBY5HRW2S","created_at":"2026-07-05T09:51:39Z"},{"alias_kind":"pith_short_8","alias_value":"VBPYWGYH","created_at":"2026-07-05T09:51:39Z"}],"graph_snapshots":[{"event_id":"sha256:fecb90bbaef830f844cd34412e850a03adf928676974692481af0d9953600d46","target":"graph","created_at":"2026-07-05T09:51:39Z","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/2412.07675/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Despite the widespread use of LLMs due to their superior performance in various tasks, their high computational costs often lead potential users to opt for the pretraining-finetuning pipeline. However, biases prevalent in manually constructed datasets can introduce spurious correlations between tokens and labels, creating so-called shortcuts and hindering the generalizability of fine-tuned models. Existing debiasing methods often rely on prior knowledge of specific dataset biases, which is challenging to acquire a priori. We propose RAZOR (Rewriting And Zero-bias Optimization Refinement), a no","authors_text":"Bardh Prenkaj, Gjergji Kasneci, Shuo Yang","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-10T17:02:58Z","title":"RAZOR: Sharpening Knowledge by Cutting Bias with Unsupervised Text Rewriting"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.07675","kind":"arxiv","version":3},"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:efb58da0370b46c2ed029eae589d51c9275f929f03db27741375a75da59aabe3","target":"record","created_at":"2026-07-05T09:51:39Z","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":"53fdad395a9eb6347cb1b03353d1ebe6e69a0ba407e323acf4fda68854661116","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-10T17:02:58Z","title_canon_sha256":"14bf03d5a9347ab02c6a56d6fdcbd5d7e63e4d6daac3b8eb2e599a03c15f93b8"},"schema_version":"1.0","source":{"id":"2412.07675","kind":"arxiv","version":3}},"canonical_sha256":"a85f8b1b070e3a78db5202843abb628380e3a9a4325b787c07c6062936dae628","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a85f8b1b070e3a78db5202843abb628380e3a9a4325b787c07c6062936dae628","first_computed_at":"2026-07-05T09:51:39.362589Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:51:39.362589Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"elVp51yYLmcwY6zs5LBsfRiepQtqVx5As+C4L3nieaOlMX3tR/S1HnqXqBRS6AL9KC2TFIIeoCBCvGr8He0rDw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:51:39.363020Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.07675","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:efb58da0370b46c2ed029eae589d51c9275f929f03db27741375a75da59aabe3","sha256:fecb90bbaef830f844cd34412e850a03adf928676974692481af0d9953600d46"],"state_sha256":"cc6aefbd0cb09c99b9eb9df57a2b9aaf800498626b5f50cdbff9b223e5b46bf9"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3llsMYHSsIbmIhwO7LI65crdM2A5dwI5xcd2LL8nNJs2RPA31zu4f98j5hskaA6HNk9csKODQfb3zuxEpHE5AQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T18:16:19.777706Z","bundle_sha256":"7d99744a2a2c7d0871866a495e9e29258beb08fc19f79008e7211157a76b3c6d"}}