{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:O4CBYOB5WUMWHBKC7EVCG2KQYK","short_pith_number":"pith:O4CBYOB5","canonical_record":{"source":{"id":"2506.15894","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T21:35:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"14f554f3853671c9e28b34a3f27057c84a9a5619ed1c76c5c12e8fea0058cda6","abstract_canon_sha256":"15f241fdde99496b1c118f1b5ab07fe469fdea417d13025c94f2c4e834b7a1dc"},"schema_version":"1.0"},"canonical_sha256":"77041c383db519638542f92a236950c29c339641819bb0273efe4ceb9060cf61","source":{"kind":"arxiv","id":"2506.15894","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.15894","created_at":"2026-07-05T11:24:26Z"},{"alias_kind":"arxiv_version","alias_value":"2506.15894v1","created_at":"2026-07-05T11:24:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.15894","created_at":"2026-07-05T11:24:26Z"},{"alias_kind":"pith_short_12","alias_value":"O4CBYOB5WUMW","created_at":"2026-07-05T11:24:26Z"},{"alias_kind":"pith_short_16","alias_value":"O4CBYOB5WUMWHBKC","created_at":"2026-07-05T11:24:26Z"},{"alias_kind":"pith_short_8","alias_value":"O4CBYOB5","created_at":"2026-07-05T11:24:26Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:O4CBYOB5WUMWHBKC7EVCG2KQYK","target":"record","payload":{"canonical_record":{"source":{"id":"2506.15894","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T21:35:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"14f554f3853671c9e28b34a3f27057c84a9a5619ed1c76c5c12e8fea0058cda6","abstract_canon_sha256":"15f241fdde99496b1c118f1b5ab07fe469fdea417d13025c94f2c4e834b7a1dc"},"schema_version":"1.0"},"canonical_sha256":"77041c383db519638542f92a236950c29c339641819bb0273efe4ceb9060cf61","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:24:26.010272Z","signature_b64":"B6v2GXTZ5b/0Chgj+P8x0d1cX1RJb7/1CERyoDcbMjJfhK/hZ/Sj7KJXXVME0C5MwqcMnc5JRYBraSrawQBQDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"77041c383db519638542f92a236950c29c339641819bb0273efe4ceb9060cf61","last_reissued_at":"2026-07-05T11:24:26.009773Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:24:26.009773Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2506.15894","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":"2026-07-05T11:24:26Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+405Ogq8BLEDYFpusvFIozNwOMR0e0ERifmI3A8JP/7Uel+TISBxWAO89Hd7kmnQmtsI+Bu7Vs+J8w+rotsBDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T04:36:50.333372Z"},"content_sha256":"84d6aab2276ef107a66c01ce9a3b66cf6957a192f6ea80d4e1d5ceb3cd8fc61f","schema_version":"1.0","event_id":"sha256:84d6aab2276ef107a66c01ce9a3b66cf6957a192f6ea80d4e1d5ceb3cd8fc61f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:O4CBYOB5WUMWHBKC7EVCG2KQYK","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Language Models can perform Single-Utterance Self-Correction of Perturbed Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Eddie Kim, Ivan Zhang, Jimin Sun, Sam Silver, Sara Hooker","submitted_at":"2025-06-18T21:35:44Z","abstract_excerpt":"Large Language Models (LLMs) have demonstrated impressive mathematical reasoning capabilities, yet their performance remains brittle to minor variations in problem description and prompting strategy. Furthermore, reasoning is vulnerable to sampling-induced errors which autoregressive models must primarily address using self-correction via additionally-generated tokens. To better understand self-correction capabilities of recent models, we conduct experiments measuring models' ability to self-correct synthetic perturbations introduced into their Chain of Thought (CoT) reasoning. We observe robu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.15894","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/2506.15894/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":"2026-07-05T11:24:26Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KAVdZoxG8bCFgLnFKnFW6615CCkrbVXvTkdQI5LZPGLzvrNGtiN4QiECLvrQxWcDDWggesKdqj/PKYtgVmUTBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T04:36:50.334023Z"},"content_sha256":"8bdbb116021c4bf96e159d10bb7713bdc58e48247b641f5f4f732fee08e76c46","schema_version":"1.0","event_id":"sha256:8bdbb116021c4bf96e159d10bb7713bdc58e48247b641f5f4f732fee08e76c46"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/O4CBYOB5WUMWHBKC7EVCG2KQYK/bundle.json","state_url":"https://pith.science/pith/O4CBYOB5WUMWHBKC7EVCG2KQYK/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/O4CBYOB5WUMWHBKC7EVCG2KQYK/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":"2026-07-07T04:36:50Z","links":{"resolver":"https://pith.science/pith/O4CBYOB5WUMWHBKC7EVCG2KQYK","bundle":"https://pith.science/pith/O4CBYOB5WUMWHBKC7EVCG2KQYK/bundle.json","state":"https://pith.science/pith/O4CBYOB5WUMWHBKC7EVCG2KQYK/state.json","well_known_bundle":"https://pith.science/.well-known/pith/O4CBYOB5WUMWHBKC7EVCG2KQYK/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:O4CBYOB5WUMWHBKC7EVCG2KQYK","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":"15f241fdde99496b1c118f1b5ab07fe469fdea417d13025c94f2c4e834b7a1dc","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T21:35:44Z","title_canon_sha256":"14f554f3853671c9e28b34a3f27057c84a9a5619ed1c76c5c12e8fea0058cda6"},"schema_version":"1.0","source":{"id":"2506.15894","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.15894","created_at":"2026-07-05T11:24:26Z"},{"alias_kind":"arxiv_version","alias_value":"2506.15894v1","created_at":"2026-07-05T11:24:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.15894","created_at":"2026-07-05T11:24:26Z"},{"alias_kind":"pith_short_12","alias_value":"O4CBYOB5WUMW","created_at":"2026-07-05T11:24:26Z"},{"alias_kind":"pith_short_16","alias_value":"O4CBYOB5WUMWHBKC","created_at":"2026-07-05T11:24:26Z"},{"alias_kind":"pith_short_8","alias_value":"O4CBYOB5","created_at":"2026-07-05T11:24:26Z"}],"graph_snapshots":[{"event_id":"sha256:8bdbb116021c4bf96e159d10bb7713bdc58e48247b641f5f4f732fee08e76c46","target":"graph","created_at":"2026-07-05T11:24:26Z","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/2506.15894/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) have demonstrated impressive mathematical reasoning capabilities, yet their performance remains brittle to minor variations in problem description and prompting strategy. Furthermore, reasoning is vulnerable to sampling-induced errors which autoregressive models must primarily address using self-correction via additionally-generated tokens. To better understand self-correction capabilities of recent models, we conduct experiments measuring models' ability to self-correct synthetic perturbations introduced into their Chain of Thought (CoT) reasoning. We observe robu","authors_text":"Eddie Kim, Ivan Zhang, Jimin Sun, Sam Silver, Sara Hooker","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T21:35:44Z","title":"Language Models can perform Single-Utterance Self-Correction of Perturbed Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.15894","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":"sha256:84d6aab2276ef107a66c01ce9a3b66cf6957a192f6ea80d4e1d5ceb3cd8fc61f","target":"record","created_at":"2026-07-05T11:24:26Z","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":"15f241fdde99496b1c118f1b5ab07fe469fdea417d13025c94f2c4e834b7a1dc","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T21:35:44Z","title_canon_sha256":"14f554f3853671c9e28b34a3f27057c84a9a5619ed1c76c5c12e8fea0058cda6"},"schema_version":"1.0","source":{"id":"2506.15894","kind":"arxiv","version":1}},"canonical_sha256":"77041c383db519638542f92a236950c29c339641819bb0273efe4ceb9060cf61","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"77041c383db519638542f92a236950c29c339641819bb0273efe4ceb9060cf61","first_computed_at":"2026-07-05T11:24:26.009773Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:24:26.009773Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"B6v2GXTZ5b/0Chgj+P8x0d1cX1RJb7/1CERyoDcbMjJfhK/hZ/Sj7KJXXVME0C5MwqcMnc5JRYBraSrawQBQDw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:24:26.010272Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.15894","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:84d6aab2276ef107a66c01ce9a3b66cf6957a192f6ea80d4e1d5ceb3cd8fc61f","sha256:8bdbb116021c4bf96e159d10bb7713bdc58e48247b641f5f4f732fee08e76c46"],"state_sha256":"309c737a3c3d105c3be7519dd908455073699ebc87f367b3e73728a20ffa018b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nc+bA4dvKkPi1h1k/EG3wPysGjj2PDNBzoVDx0BF4UnSTzdWFlimPFzAk/KYJlatsmPEE7aIUrtQooIMojSWBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T04:36:50.337330Z","bundle_sha256":"5afbb34e30b828aac1db9d911c4ff170bc5263a25c527ba494beccfda32af326"}}