{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:GU6GNBV2OFLXUATPJPNH46GUIC","short_pith_number":"pith:GU6GNBV2","canonical_record":{"source":{"id":"2302.14003","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-27T17:47:53Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"fd3c8b3f3985d3085942aceef4bbeb99ba5cb27ec523b36e3243f339bc16f4ca","abstract_canon_sha256":"8be9832b3a165c6612b89ff3c5f4ecc2560ec11cb2fdaf64d15fc8c1dd693c35"},"schema_version":"1.0"},"canonical_sha256":"353c6686ba71577a026f4bda7e78d440af31d4e8218ecc9dd80deee8b387a716","source":{"kind":"arxiv","id":"2302.14003","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2302.14003","created_at":"2026-07-05T05:45:53Z"},{"alias_kind":"arxiv_version","alias_value":"2302.14003v1","created_at":"2026-07-05T05:45:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.14003","created_at":"2026-07-05T05:45:53Z"},{"alias_kind":"pith_short_12","alias_value":"GU6GNBV2OFLX","created_at":"2026-07-05T05:45:53Z"},{"alias_kind":"pith_short_16","alias_value":"GU6GNBV2OFLXUATP","created_at":"2026-07-05T05:45:53Z"},{"alias_kind":"pith_short_8","alias_value":"GU6GNBV2","created_at":"2026-07-05T05:45:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:GU6GNBV2OFLXUATPJPNH46GUIC","target":"record","payload":{"canonical_record":{"source":{"id":"2302.14003","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-27T17:47:53Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"fd3c8b3f3985d3085942aceef4bbeb99ba5cb27ec523b36e3243f339bc16f4ca","abstract_canon_sha256":"8be9832b3a165c6612b89ff3c5f4ecc2560ec11cb2fdaf64d15fc8c1dd693c35"},"schema_version":"1.0"},"canonical_sha256":"353c6686ba71577a026f4bda7e78d440af31d4e8218ecc9dd80deee8b387a716","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:45:53.336108Z","signature_b64":"Wn2oWye2blQNW86BotPE4rw+QtCH+EbVt1jg/9wtOSfdTLZaTmsNTGhXHx7Jr5uBflaGY1X8tnzdIM+HfjpSDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"353c6686ba71577a026f4bda7e78d440af31d4e8218ecc9dd80deee8b387a716","last_reissued_at":"2026-07-05T05:45:53.335615Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:45:53.335615Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2302.14003","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":"2026-07-05T05:45:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tbAZwP5L6a5JMh6LLLvAZ040btrYEqESkQnAByF1ypNXONqKos5pMGdhtxmV9Qsx2azhmuekVnWRybJBl+yCDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T11:51:22.477924Z"},"content_sha256":"aee3168913189c7b9687c4f86aaf53dd5542ff9c0587352ab093df0050ddf804","schema_version":"1.0","event_id":"sha256:aee3168913189c7b9687c4f86aaf53dd5542ff9c0587352ab093df0050ddf804"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:GU6GNBV2OFLXUATPJPNH46GUIC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Systematic Rectification of Language Models via Dead-end Analysis","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Jackie Chi Kit Cheung, Mehdi Fatemi, Meng Cao, Samira Shabanian","submitted_at":"2023-02-27T17:47:53Z","abstract_excerpt":"With adversarial or otherwise normal prompts, existing large language models (LLM) can be pushed to generate toxic discourses. One way to reduce the risk of LLMs generating undesired discourses is to alter the training of the LLM. This can be very restrictive due to demanding computation requirements. Other methods rely on rule-based or prompt-based token elimination, which are limited as they dismiss future tokens and the overall meaning of the complete discourse. Here, we center detoxification on the probability that the finished discourse is ultimately considered toxic. That is, at each poi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2302.14003","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/2302.14003/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":"2026-07-05T05:45:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZydIh8hgaE5XebiGJk9IgfA7QGQYcQUlihDdWZfwaYU18oodNMtvRSQIctIVGdPU7fADqxv7lfQ46bCdReadDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T11:51:22.478537Z"},"content_sha256":"ec744b051d9faceefd659e8f6d1027f92581215fa1726cf972f62e64f57e67c5","schema_version":"1.0","event_id":"sha256:ec744b051d9faceefd659e8f6d1027f92581215fa1726cf972f62e64f57e67c5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/GU6GNBV2OFLXUATPJPNH46GUIC/bundle.json","state_url":"https://pith.science/pith/GU6GNBV2OFLXUATPJPNH46GUIC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/GU6GNBV2OFLXUATPJPNH46GUIC/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":"2026-08-08T11:51:22Z","links":{"resolver":"https://pith.science/pith/GU6GNBV2OFLXUATPJPNH46GUIC","bundle":"https://pith.science/pith/GU6GNBV2OFLXUATPJPNH46GUIC/bundle.json","state":"https://pith.science/pith/GU6GNBV2OFLXUATPJPNH46GUIC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/GU6GNBV2OFLXUATPJPNH46GUIC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:GU6GNBV2OFLXUATPJPNH46GUIC","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":"8be9832b3a165c6612b89ff3c5f4ecc2560ec11cb2fdaf64d15fc8c1dd693c35","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-27T17:47:53Z","title_canon_sha256":"fd3c8b3f3985d3085942aceef4bbeb99ba5cb27ec523b36e3243f339bc16f4ca"},"schema_version":"1.0","source":{"id":"2302.14003","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2302.14003","created_at":"2026-07-05T05:45:53Z"},{"alias_kind":"arxiv_version","alias_value":"2302.14003v1","created_at":"2026-07-05T05:45:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.14003","created_at":"2026-07-05T05:45:53Z"},{"alias_kind":"pith_short_12","alias_value":"GU6GNBV2OFLX","created_at":"2026-07-05T05:45:53Z"},{"alias_kind":"pith_short_16","alias_value":"GU6GNBV2OFLXUATP","created_at":"2026-07-05T05:45:53Z"},{"alias_kind":"pith_short_8","alias_value":"GU6GNBV2","created_at":"2026-07-05T05:45:53Z"}],"graph_snapshots":[{"event_id":"sha256:ec744b051d9faceefd659e8f6d1027f92581215fa1726cf972f62e64f57e67c5","target":"graph","created_at":"2026-07-05T05:45:53Z","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/2302.14003/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"With adversarial or otherwise normal prompts, existing large language models (LLM) can be pushed to generate toxic discourses. One way to reduce the risk of LLMs generating undesired discourses is to alter the training of the LLM. This can be very restrictive due to demanding computation requirements. Other methods rely on rule-based or prompt-based token elimination, which are limited as they dismiss future tokens and the overall meaning of the complete discourse. Here, we center detoxification on the probability that the finished discourse is ultimately considered toxic. That is, at each poi","authors_text":"Jackie Chi Kit Cheung, Mehdi Fatemi, Meng Cao, Samira Shabanian","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-27T17:47:53Z","title":"Systematic Rectification of Language Models via Dead-end Analysis"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2302.14003","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:aee3168913189c7b9687c4f86aaf53dd5542ff9c0587352ab093df0050ddf804","target":"record","created_at":"2026-07-05T05:45:53Z","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":"8be9832b3a165c6612b89ff3c5f4ecc2560ec11cb2fdaf64d15fc8c1dd693c35","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-27T17:47:53Z","title_canon_sha256":"fd3c8b3f3985d3085942aceef4bbeb99ba5cb27ec523b36e3243f339bc16f4ca"},"schema_version":"1.0","source":{"id":"2302.14003","kind":"arxiv","version":1}},"canonical_sha256":"353c6686ba71577a026f4bda7e78d440af31d4e8218ecc9dd80deee8b387a716","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"353c6686ba71577a026f4bda7e78d440af31d4e8218ecc9dd80deee8b387a716","first_computed_at":"2026-07-05T05:45:53.335615Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:45:53.335615Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Wn2oWye2blQNW86BotPE4rw+QtCH+EbVt1jg/9wtOSfdTLZaTmsNTGhXHx7Jr5uBflaGY1X8tnzdIM+HfjpSDA==","signature_status":"signed_v1","signed_at":"2026-07-05T05:45:53.336108Z","signed_message":"canonical_sha256_bytes"},"source_id":"2302.14003","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:aee3168913189c7b9687c4f86aaf53dd5542ff9c0587352ab093df0050ddf804","sha256:ec744b051d9faceefd659e8f6d1027f92581215fa1726cf972f62e64f57e67c5"],"state_sha256":"6a0c0d30bfde061a408a2c79b8206c46c5199e63f888e0968ff7cb0e6eae68bf"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uhHqucFyhTTc2Nye7Yv9Rc9tOBeZg97g5jGd2ovCxBeuRQ/e9nEbZ2T3wRUd90KXFUoyWGIUVbcFjFbToVcGDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T11:51:22.482311Z","bundle_sha256":"59e69ee093be868dcf3a317c97dfba6f7f536ff9c45dfcaf91f017e19ed06d62"}}