{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:ZDUEQQ3KCC4H2H2Z5IENLOAQ4F","short_pith_number":"pith:ZDUEQQ3K","canonical_record":{"source":{"id":"2212.09721","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-12-19T18:49:09Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"245b2be149bc7f1323ef98f17814b2991d8d221693e978c92a66dd982d6a98e1","abstract_canon_sha256":"9bf9fef6ff2a160cf30e4cee124b959d8534b17ad1267350d4a205a97fee1962"},"schema_version":"1.0"},"canonical_sha256":"c8e848436a10b87d1f59ea08d5b810e162503335fd14b800d7294ec566ac5b61","source":{"kind":"arxiv","id":"2212.09721","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2212.09721","created_at":"2026-07-05T06:11:59Z"},{"alias_kind":"arxiv_version","alias_value":"2212.09721v2","created_at":"2026-07-05T06:11:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2212.09721","created_at":"2026-07-05T06:11:59Z"},{"alias_kind":"pith_short_12","alias_value":"ZDUEQQ3KCC4H","created_at":"2026-07-05T06:11:59Z"},{"alias_kind":"pith_short_16","alias_value":"ZDUEQQ3KCC4H2H2Z","created_at":"2026-07-05T06:11:59Z"},{"alias_kind":"pith_short_8","alias_value":"ZDUEQQ3K","created_at":"2026-07-05T06:11:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:ZDUEQQ3KCC4H2H2Z5IENLOAQ4F","target":"record","payload":{"canonical_record":{"source":{"id":"2212.09721","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-12-19T18:49:09Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"245b2be149bc7f1323ef98f17814b2991d8d221693e978c92a66dd982d6a98e1","abstract_canon_sha256":"9bf9fef6ff2a160cf30e4cee124b959d8534b17ad1267350d4a205a97fee1962"},"schema_version":"1.0"},"canonical_sha256":"c8e848436a10b87d1f59ea08d5b810e162503335fd14b800d7294ec566ac5b61","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:11:59.094121Z","signature_b64":"fRaOhmh/1ZlLa8XmXRwT5tExlovIAsc2FChir6T85oqXWEZOmIXmtUGAJz3xf2hTG3TW45rc0UolCSnnKt2ZDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c8e848436a10b87d1f59ea08d5b810e162503335fd14b800d7294ec566ac5b61","last_reissued_at":"2026-07-05T06:11:59.093724Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:11:59.093724Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2212.09721","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":"2026-07-05T06:11:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PYgtdlKVpb70FMCkUW4v9JvNn5nSfWVwzxrL7KEheeqKoSzAS2tRqA6m1EUxGZldxe8zCM/HvvxIyKlQQgR0AA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T06:01:52.896826Z"},"content_sha256":"959142c3a44ed9b4fd41355cc11859097b19c4e47b9bf53a3e9807262667301d","schema_version":"1.0","event_id":"sha256:959142c3a44ed9b4fd41355cc11859097b19c4e47b9bf53a3e9807262667301d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:ZDUEQQ3KCC4H2H2Z5IENLOAQ4F","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"KNIFE: Distilling Reasoning Knowledge From Free-Text Rationales","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Aaron Chan, Brihi Joshi, Hanjie Chen, Wyatt Lake, Xiang Ren, Zhiyuan Zeng","submitted_at":"2022-12-19T18:49:09Z","abstract_excerpt":"Language models (LMs) have yielded impressive results on many language reasoning tasks, but their unexpected errors raise doubts about their reasoning abilities. In light of this, there is growing interest in finetuning/prompting LMs with both task instances and their associated free-text rationales (FTRs), which explain the correct reasoning process for predicting the correct task output (i.e., how to be \"right for the right reasons\"). However, existing finetuning methods fail to improve LM performance, while prompting needs prohibitively large (i.e., >50B) LMs to work well. We propose KNIFE,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2212.09721","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/2212.09721/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":"2026-07-05T06:11:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lVgqnRkjraZuTAN3uwRoaIuEAcDVumYVrSRrtiQDG8UKwAbo2woAZLXhso8K6HpZ4dHNrd02hsBcIrfI985LCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T06:01:52.897207Z"},"content_sha256":"e24afdcd8d924d4c6165785897cfa487a702e184a5eda7c5dbba5b59ca9d8395","schema_version":"1.0","event_id":"sha256:e24afdcd8d924d4c6165785897cfa487a702e184a5eda7c5dbba5b59ca9d8395"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZDUEQQ3KCC4H2H2Z5IENLOAQ4F/bundle.json","state_url":"https://pith.science/pith/ZDUEQQ3KCC4H2H2Z5IENLOAQ4F/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZDUEQQ3KCC4H2H2Z5IENLOAQ4F/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":"2026-07-07T06:01:52Z","links":{"resolver":"https://pith.science/pith/ZDUEQQ3KCC4H2H2Z5IENLOAQ4F","bundle":"https://pith.science/pith/ZDUEQQ3KCC4H2H2Z5IENLOAQ4F/bundle.json","state":"https://pith.science/pith/ZDUEQQ3KCC4H2H2Z5IENLOAQ4F/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZDUEQQ3KCC4H2H2Z5IENLOAQ4F/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:ZDUEQQ3KCC4H2H2Z5IENLOAQ4F","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":"9bf9fef6ff2a160cf30e4cee124b959d8534b17ad1267350d4a205a97fee1962","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-12-19T18:49:09Z","title_canon_sha256":"245b2be149bc7f1323ef98f17814b2991d8d221693e978c92a66dd982d6a98e1"},"schema_version":"1.0","source":{"id":"2212.09721","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2212.09721","created_at":"2026-07-05T06:11:59Z"},{"alias_kind":"arxiv_version","alias_value":"2212.09721v2","created_at":"2026-07-05T06:11:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2212.09721","created_at":"2026-07-05T06:11:59Z"},{"alias_kind":"pith_short_12","alias_value":"ZDUEQQ3KCC4H","created_at":"2026-07-05T06:11:59Z"},{"alias_kind":"pith_short_16","alias_value":"ZDUEQQ3KCC4H2H2Z","created_at":"2026-07-05T06:11:59Z"},{"alias_kind":"pith_short_8","alias_value":"ZDUEQQ3K","created_at":"2026-07-05T06:11:59Z"}],"graph_snapshots":[{"event_id":"sha256:e24afdcd8d924d4c6165785897cfa487a702e184a5eda7c5dbba5b59ca9d8395","target":"graph","created_at":"2026-07-05T06:11:59Z","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/2212.09721/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Language models (LMs) have yielded impressive results on many language reasoning tasks, but their unexpected errors raise doubts about their reasoning abilities. In light of this, there is growing interest in finetuning/prompting LMs with both task instances and their associated free-text rationales (FTRs), which explain the correct reasoning process for predicting the correct task output (i.e., how to be \"right for the right reasons\"). However, existing finetuning methods fail to improve LM performance, while prompting needs prohibitively large (i.e., >50B) LMs to work well. We propose KNIFE,","authors_text":"Aaron Chan, Brihi Joshi, Hanjie Chen, Wyatt Lake, Xiang Ren, Zhiyuan Zeng","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-12-19T18:49:09Z","title":"KNIFE: Distilling Reasoning Knowledge From Free-Text Rationales"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2212.09721","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":"sha256:959142c3a44ed9b4fd41355cc11859097b19c4e47b9bf53a3e9807262667301d","target":"record","created_at":"2026-07-05T06:11:59Z","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":"9bf9fef6ff2a160cf30e4cee124b959d8534b17ad1267350d4a205a97fee1962","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-12-19T18:49:09Z","title_canon_sha256":"245b2be149bc7f1323ef98f17814b2991d8d221693e978c92a66dd982d6a98e1"},"schema_version":"1.0","source":{"id":"2212.09721","kind":"arxiv","version":2}},"canonical_sha256":"c8e848436a10b87d1f59ea08d5b810e162503335fd14b800d7294ec566ac5b61","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c8e848436a10b87d1f59ea08d5b810e162503335fd14b800d7294ec566ac5b61","first_computed_at":"2026-07-05T06:11:59.093724Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:11:59.093724Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"fRaOhmh/1ZlLa8XmXRwT5tExlovIAsc2FChir6T85oqXWEZOmIXmtUGAJz3xf2hTG3TW45rc0UolCSnnKt2ZDg==","signature_status":"signed_v1","signed_at":"2026-07-05T06:11:59.094121Z","signed_message":"canonical_sha256_bytes"},"source_id":"2212.09721","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:959142c3a44ed9b4fd41355cc11859097b19c4e47b9bf53a3e9807262667301d","sha256:e24afdcd8d924d4c6165785897cfa487a702e184a5eda7c5dbba5b59ca9d8395"],"state_sha256":"1338893dcebe28ab126a10e7d89a931bc043ddcd9afd65206f71497099ad0833"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2lDN8SDxU+DV4yGv8XYShjoySUEGsBtIN5LsSx/TR6OYN3WO0ek3LrEo0whEMG+UujR7D7GanR8vhUe1HADwDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T06:01:52.899440Z","bundle_sha256":"197dd794d6a15f0bc59cf6a0c8a2d80824ff7b414ac21d7f9728793f264d8ff7"}}