{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:P2D4BI7FRYWSEHFJJRNCJWZ4ZK","short_pith_number":"pith:P2D4BI7F","canonical_record":{"source":{"id":"2412.14959","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-19T15:39:31Z","cross_cats_sorted":[],"title_canon_sha256":"dbda741a3a58c88c040fd7ab69db8eca213353d02ae1c79f6c1fe7eaa47f9c81","abstract_canon_sha256":"a890ee9650f187475c06740dc8875156a98bc0af6ffe8e54fc96d7cfbd51fff3"},"schema_version":"1.0"},"canonical_sha256":"7e87c0a3e58e2d221ca94c5a24db3cca8f2db1c674c8529afe602956bc7abf02","source":{"kind":"arxiv","id":"2412.14959","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.14959","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"arxiv_version","alias_value":"2412.14959v2","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.14959","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_12","alias_value":"P2D4BI7FRYWS","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_16","alias_value":"P2D4BI7FRYWSEHFJ","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_8","alias_value":"P2D4BI7F","created_at":"2026-07-05T11:37:14Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:P2D4BI7FRYWSEHFJJRNCJWZ4ZK","target":"record","payload":{"canonical_record":{"source":{"id":"2412.14959","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-19T15:39:31Z","cross_cats_sorted":[],"title_canon_sha256":"dbda741a3a58c88c040fd7ab69db8eca213353d02ae1c79f6c1fe7eaa47f9c81","abstract_canon_sha256":"a890ee9650f187475c06740dc8875156a98bc0af6ffe8e54fc96d7cfbd51fff3"},"schema_version":"1.0"},"canonical_sha256":"7e87c0a3e58e2d221ca94c5a24db3cca8f2db1c674c8529afe602956bc7abf02","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:37:14.006075Z","signature_b64":"jLX8gbtVarUAr3tv1cHjeO0o1AbtrScXKjTNJIiRiEdRMyArsasFmEV1okgFHqrQmGcfZ3EDUTyM10iS0EmXBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7e87c0a3e58e2d221ca94c5a24db3cca8f2db1c674c8529afe602956bc7abf02","last_reissued_at":"2026-07-05T11:37:14.005585Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:37:14.005585Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.14959","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":"2026-07-05T11:37:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Mmqgvo84aEUoVFNS3vXzBO3eIkmKSLm35Wf2ckg20gqJtsQa5OZdlYEM46Rm5QJWzbxArFTKJwhbXmR2UEzaBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T08:59:30.913571Z"},"content_sha256":"ba0885cd2fb5bd1fa674b9b9feaaf441e0f925d2c9c2454fba36f0694abdc60e","schema_version":"1.0","event_id":"sha256:ba0885cd2fb5bd1fa674b9b9feaaf441e0f925d2c9c2454fba36f0694abdc60e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:P2D4BI7FRYWSEHFJJRNCJWZ4ZK","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Understanding the Dark Side of LLMs' Intrinsic Self-Correction","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Di Wang, Han Qiu, Haoting Qian, Hewu Li, Ke Xu, Minlie Huang, Qingjie Zhang, Tianwei Zhang, Yan Liu, Yiming Li","submitted_at":"2024-12-19T15:39:31Z","abstract_excerpt":"Intrinsic self-correction was proposed to improve LLMs' responses via feedback prompts solely based on their inherent capability. However, recent works show that LLMs' intrinsic self-correction fails without oracle labels as feedback prompts. In this paper, we aim to interpret LLMs' intrinsic self-correction for different tasks, especially for those failure cases. By including one simple task and three complex tasks with state-of-the-art (SOTA) LLMs like ChatGPT families (o1, 4o, 3.5-turbo) and Llama families (2-7B, 3-8B, and 3.1-8B), we design three interpretation methods to reveal the dark s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.14959","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/2412.14959/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":"2026-07-05T11:37:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Vs3vywjHXZ5Jl7pO022dE1F7MK/7r4W1e4iSPerQj0D7cZ2Rq7/K0OYci2/Yg9imi5tSBkmIkBhjwWgCw096Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T08:59:30.914084Z"},"content_sha256":"d90871d04e62c9961a5e0853afe932d2146c105b390f0f432ede402b53114fb3","schema_version":"1.0","event_id":"sha256:d90871d04e62c9961a5e0853afe932d2146c105b390f0f432ede402b53114fb3"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/P2D4BI7FRYWSEHFJJRNCJWZ4ZK/bundle.json","state_url":"https://pith.science/pith/P2D4BI7FRYWSEHFJJRNCJWZ4ZK/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/P2D4BI7FRYWSEHFJJRNCJWZ4ZK/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":"2026-08-19T08:59:30Z","links":{"resolver":"https://pith.science/pith/P2D4BI7FRYWSEHFJJRNCJWZ4ZK","bundle":"https://pith.science/pith/P2D4BI7FRYWSEHFJJRNCJWZ4ZK/bundle.json","state":"https://pith.science/pith/P2D4BI7FRYWSEHFJJRNCJWZ4ZK/state.json","well_known_bundle":"https://pith.science/.well-known/pith/P2D4BI7FRYWSEHFJJRNCJWZ4ZK/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:P2D4BI7FRYWSEHFJJRNCJWZ4ZK","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":"a890ee9650f187475c06740dc8875156a98bc0af6ffe8e54fc96d7cfbd51fff3","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-19T15:39:31Z","title_canon_sha256":"dbda741a3a58c88c040fd7ab69db8eca213353d02ae1c79f6c1fe7eaa47f9c81"},"schema_version":"1.0","source":{"id":"2412.14959","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.14959","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"arxiv_version","alias_value":"2412.14959v2","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.14959","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_12","alias_value":"P2D4BI7FRYWS","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_16","alias_value":"P2D4BI7FRYWSEHFJ","created_at":"2026-07-05T11:37:14Z"},{"alias_kind":"pith_short_8","alias_value":"P2D4BI7F","created_at":"2026-07-05T11:37:14Z"}],"graph_snapshots":[{"event_id":"sha256:d90871d04e62c9961a5e0853afe932d2146c105b390f0f432ede402b53114fb3","target":"graph","created_at":"2026-07-05T11:37:14Z","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/2412.14959/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Intrinsic self-correction was proposed to improve LLMs' responses via feedback prompts solely based on their inherent capability. However, recent works show that LLMs' intrinsic self-correction fails without oracle labels as feedback prompts. In this paper, we aim to interpret LLMs' intrinsic self-correction for different tasks, especially for those failure cases. By including one simple task and three complex tasks with state-of-the-art (SOTA) LLMs like ChatGPT families (o1, 4o, 3.5-turbo) and Llama families (2-7B, 3-8B, and 3.1-8B), we design three interpretation methods to reveal the dark s","authors_text":"Di Wang, Han Qiu, Haoting Qian, Hewu Li, Ke Xu, Minlie Huang, Qingjie Zhang, Tianwei Zhang, Yan Liu, Yiming Li","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-19T15:39:31Z","title":"Understanding the Dark Side of LLMs' Intrinsic Self-Correction"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.14959","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:ba0885cd2fb5bd1fa674b9b9feaaf441e0f925d2c9c2454fba36f0694abdc60e","target":"record","created_at":"2026-07-05T11:37:14Z","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":"a890ee9650f187475c06740dc8875156a98bc0af6ffe8e54fc96d7cfbd51fff3","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-19T15:39:31Z","title_canon_sha256":"dbda741a3a58c88c040fd7ab69db8eca213353d02ae1c79f6c1fe7eaa47f9c81"},"schema_version":"1.0","source":{"id":"2412.14959","kind":"arxiv","version":2}},"canonical_sha256":"7e87c0a3e58e2d221ca94c5a24db3cca8f2db1c674c8529afe602956bc7abf02","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7e87c0a3e58e2d221ca94c5a24db3cca8f2db1c674c8529afe602956bc7abf02","first_computed_at":"2026-07-05T11:37:14.005585Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:37:14.005585Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"jLX8gbtVarUAr3tv1cHjeO0o1AbtrScXKjTNJIiRiEdRMyArsasFmEV1okgFHqrQmGcfZ3EDUTyM10iS0EmXBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:37:14.006075Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.14959","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ba0885cd2fb5bd1fa674b9b9feaaf441e0f925d2c9c2454fba36f0694abdc60e","sha256:d90871d04e62c9961a5e0853afe932d2146c105b390f0f432ede402b53114fb3"],"state_sha256":"647f515360e8ece8ba15ca330d588e5085f6ea9a1f35ce257818775862db1e45"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TlPEplhsbogNAuiCiL7ftvp16PVa0wFwgGk+DPX3ThrR2zu/OpPDLcB9quwFM/+si0Ztrcn5IQL2furldvFJDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-19T08:59:30.920338Z","bundle_sha256":"248670f35efeb781bcc753a6585ed2adbce6ae5c8cca925a0e0bfeaa98fee798"}}