{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:6HZGICBEC4BDLRXZ53YLCRDFYU","short_pith_number":"pith:6HZGICBE","canonical_record":{"source":{"id":"2411.12768","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-18T07:52:12Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"40688d25904492a540bbc3396519f1f3a477b406fb512071c5cd7eb27fd4f626","abstract_canon_sha256":"47569937c4096e6e98f0063010dc1002ca8e8d4f6bd7458f3f461e1ad72e185f"},"schema_version":"1.0"},"canonical_sha256":"f1f2640824170235c6f9eef0b14465c535e3d0063c2291ded20deaffdf51854c","source":{"kind":"arxiv","id":"2411.12768","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.12768","created_at":"2026-07-05T11:19:50Z"},{"alias_kind":"arxiv_version","alias_value":"2411.12768v2","created_at":"2026-07-05T11:19:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.12768","created_at":"2026-07-05T11:19:50Z"},{"alias_kind":"pith_short_12","alias_value":"6HZGICBEC4BD","created_at":"2026-07-05T11:19:50Z"},{"alias_kind":"pith_short_16","alias_value":"6HZGICBEC4BDLRXZ","created_at":"2026-07-05T11:19:50Z"},{"alias_kind":"pith_short_8","alias_value":"6HZGICBE","created_at":"2026-07-05T11:19:50Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:6HZGICBEC4BDLRXZ53YLCRDFYU","target":"record","payload":{"canonical_record":{"source":{"id":"2411.12768","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-18T07:52:12Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"40688d25904492a540bbc3396519f1f3a477b406fb512071c5cd7eb27fd4f626","abstract_canon_sha256":"47569937c4096e6e98f0063010dc1002ca8e8d4f6bd7458f3f461e1ad72e185f"},"schema_version":"1.0"},"canonical_sha256":"f1f2640824170235c6f9eef0b14465c535e3d0063c2291ded20deaffdf51854c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:19:50.084226Z","signature_b64":"YmDY2TyVad7CzVUtanMai3rAJgrNkbRsaGbdkddt/66NHGsPWzXqBPQkak5EmcPu0UqDbO9EVP1kmVacJpz6BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f1f2640824170235c6f9eef0b14465c535e3d0063c2291ded20deaffdf51854c","last_reissued_at":"2026-07-05T11:19:50.083812Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:19:50.083812Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2411.12768","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":"2026-07-05T11:19:50Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ROG5M6/v4YEv2EkUhQftaoEE1Dkram3R9+2XuQOiLWANmyg0AMfTEARm2rLiP7NiOcuCv74cZit0bJw1WvisAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T20:59:06.806468Z"},"content_sha256":"2c3422844d69f90f91a08a447b88db575215959335a88a88df9ae9d6c0e49165","schema_version":"1.0","event_id":"sha256:2c3422844d69f90f91a08a447b88db575215959335a88a88df9ae9d6c0e49165"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:6HZGICBEC4BDLRXZ53YLCRDFYU","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"CROW: Eliminating Backdoors from Large Language Models via Internal Consistency Regularization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Jun Sun, Long H. Pham, Nay Myat Min, Yige Li","submitted_at":"2024-11-18T07:52:12Z","abstract_excerpt":"Large Language Models (LLMs) are vulnerable to backdoor attacks that manipulate outputs via hidden triggers. Existing defense methods--designed for vision/text classification tasks--fail for text generation. We propose Internal Consistency Regularization (CROW), a defense leveraging the observation that backdoored models exhibit unstable layer-wise hidden representations when triggered, while clean models show smooth transitions. CROW enforces consistency across layers via adversarial perturbations and regularization during finetuning, neutralizing backdoors without requiring clean reference m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.12768","kind":"arxiv","version":2},"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/2411.12768/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":"2026-07-05T11:19:50Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jm+R8vFzfnQjRr+kxLDkC6B2Rhkt2H0oFQSplqszq2smaqbhift+nkVDrgDW4uopvmZfqxKT/S+xf/rlCV3vAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T20:59:06.806950Z"},"content_sha256":"8aef62d30b46b17fa4280d9a1ec9732357749a2ff2f5abc50ddfe34d4497524b","schema_version":"1.0","event_id":"sha256:8aef62d30b46b17fa4280d9a1ec9732357749a2ff2f5abc50ddfe34d4497524b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6HZGICBEC4BDLRXZ53YLCRDFYU/bundle.json","state_url":"https://pith.science/pith/6HZGICBEC4BDLRXZ53YLCRDFYU/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6HZGICBEC4BDLRXZ53YLCRDFYU/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":"2026-08-07T20:59:06Z","links":{"resolver":"https://pith.science/pith/6HZGICBEC4BDLRXZ53YLCRDFYU","bundle":"https://pith.science/pith/6HZGICBEC4BDLRXZ53YLCRDFYU/bundle.json","state":"https://pith.science/pith/6HZGICBEC4BDLRXZ53YLCRDFYU/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6HZGICBEC4BDLRXZ53YLCRDFYU/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:6HZGICBEC4BDLRXZ53YLCRDFYU","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":"47569937c4096e6e98f0063010dc1002ca8e8d4f6bd7458f3f461e1ad72e185f","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-18T07:52:12Z","title_canon_sha256":"40688d25904492a540bbc3396519f1f3a477b406fb512071c5cd7eb27fd4f626"},"schema_version":"1.0","source":{"id":"2411.12768","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.12768","created_at":"2026-07-05T11:19:50Z"},{"alias_kind":"arxiv_version","alias_value":"2411.12768v2","created_at":"2026-07-05T11:19:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.12768","created_at":"2026-07-05T11:19:50Z"},{"alias_kind":"pith_short_12","alias_value":"6HZGICBEC4BD","created_at":"2026-07-05T11:19:50Z"},{"alias_kind":"pith_short_16","alias_value":"6HZGICBEC4BDLRXZ","created_at":"2026-07-05T11:19:50Z"},{"alias_kind":"pith_short_8","alias_value":"6HZGICBE","created_at":"2026-07-05T11:19:50Z"}],"graph_snapshots":[{"event_id":"sha256:8aef62d30b46b17fa4280d9a1ec9732357749a2ff2f5abc50ddfe34d4497524b","target":"graph","created_at":"2026-07-05T11:19:50Z","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/2411.12768/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) are vulnerable to backdoor attacks that manipulate outputs via hidden triggers. Existing defense methods--designed for vision/text classification tasks--fail for text generation. We propose Internal Consistency Regularization (CROW), a defense leveraging the observation that backdoored models exhibit unstable layer-wise hidden representations when triggered, while clean models show smooth transitions. CROW enforces consistency across layers via adversarial perturbations and regularization during finetuning, neutralizing backdoors without requiring clean reference m","authors_text":"Jun Sun, Long H. Pham, Nay Myat Min, Yige Li","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-18T07:52:12Z","title":"CROW: Eliminating Backdoors from Large Language Models via Internal Consistency Regularization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.12768","kind":"arxiv","version":2},"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":"sha256:2c3422844d69f90f91a08a447b88db575215959335a88a88df9ae9d6c0e49165","target":"record","created_at":"2026-07-05T11:19:50Z","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":"47569937c4096e6e98f0063010dc1002ca8e8d4f6bd7458f3f461e1ad72e185f","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-18T07:52:12Z","title_canon_sha256":"40688d25904492a540bbc3396519f1f3a477b406fb512071c5cd7eb27fd4f626"},"schema_version":"1.0","source":{"id":"2411.12768","kind":"arxiv","version":2}},"canonical_sha256":"f1f2640824170235c6f9eef0b14465c535e3d0063c2291ded20deaffdf51854c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f1f2640824170235c6f9eef0b14465c535e3d0063c2291ded20deaffdf51854c","first_computed_at":"2026-07-05T11:19:50.083812Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:19:50.083812Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"YmDY2TyVad7CzVUtanMai3rAJgrNkbRsaGbdkddt/66NHGsPWzXqBPQkak5EmcPu0UqDbO9EVP1kmVacJpz6BA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:19:50.084226Z","signed_message":"canonical_sha256_bytes"},"source_id":"2411.12768","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2c3422844d69f90f91a08a447b88db575215959335a88a88df9ae9d6c0e49165","sha256:8aef62d30b46b17fa4280d9a1ec9732357749a2ff2f5abc50ddfe34d4497524b"],"state_sha256":"ed6e4394f294a56c92e308a39eb0c4b1dcaf27fd6c6ca8b142366b6c2020584d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lfUISaJcE074Uuy3DnH6UUEiaA5N59+68g2GWaSyKugloKbxrUk97jsEVtxv/sX8UiTFoyQk++3WZz9Xd0p5BQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T20:59:06.811270Z","bundle_sha256":"25940485cc4aa055c65a976e88db41a940564b3e5642c660e2e2acfe30b0b0dd"}}