{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:BJH4GYKILODCCATMW2OHGXSA4G","short_pith_number":"pith:BJH4GYKI","canonical_record":{"source":{"id":"2608.04355","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T01:55:33Z","cross_cats_sorted":[],"title_canon_sha256":"67bb7e2dc1a1c9b40f1894b1c3a6294f2c25443dfa36c1a7587a0513aa38a115","abstract_canon_sha256":"bcd59d1670722a85d00516e6fff82551de99cd310d63d47bdbdfccd4b7cdd050"},"schema_version":"1.0"},"canonical_sha256":"0a4fc361485b8621026cb69c735e40e1a96dd5e347f010fdc0322e4cdce11017","source":{"kind":"arxiv","id":"2608.04355","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.04355","created_at":"2026-08-06T01:33:35Z"},{"alias_kind":"arxiv_version","alias_value":"2608.04355v1","created_at":"2026-08-06T01:33:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04355","created_at":"2026-08-06T01:33:35Z"},{"alias_kind":"pith_short_12","alias_value":"BJH4GYKILODC","created_at":"2026-08-06T01:33:35Z"},{"alias_kind":"pith_short_16","alias_value":"BJH4GYKILODCCATM","created_at":"2026-08-06T01:33:35Z"},{"alias_kind":"pith_short_8","alias_value":"BJH4GYKI","created_at":"2026-08-06T01:33:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:BJH4GYKILODCCATMW2OHGXSA4G","target":"record","payload":{"canonical_record":{"source":{"id":"2608.04355","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T01:55:33Z","cross_cats_sorted":[],"title_canon_sha256":"67bb7e2dc1a1c9b40f1894b1c3a6294f2c25443dfa36c1a7587a0513aa38a115","abstract_canon_sha256":"bcd59d1670722a85d00516e6fff82551de99cd310d63d47bdbdfccd4b7cdd050"},"schema_version":"1.0"},"canonical_sha256":"0a4fc361485b8621026cb69c735e40e1a96dd5e347f010fdc0322e4cdce11017","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-06T01:33:35.048700Z","signature_b64":"kIzHrTZCePGmSnalIHD95uK1o42V7oh/bkB6As+xGyGFTZFhJEhLa7cqLrAog3xpS30qYUNNz81y5Nd+xHZdCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0a4fc361485b8621026cb69c735e40e1a96dd5e347f010fdc0322e4cdce11017","last_reissued_at":"2026-08-06T01:33:35.047211Z","signature_status":"signed_v1","first_computed_at":"2026-08-06T01:33:35.047211Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.04355","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":"2026-08-06T01:33:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2jLt4Z9ViF71AFoB7iPmkcuQA/4ltVfY+B8vUi1pxYZ77U556kG1sujUYUAGvL/mp46moEzzasA0jMsUO5mFDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T02:21:02.305671Z"},"content_sha256":"711fee77ed32c8334272391086b58298dc475f0197c180ae81cc4a1bf830c924","schema_version":"1.0","event_id":"sha256:711fee77ed32c8334272391086b58298dc475f0197c180ae81cc4a1bf830c924"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:BJH4GYKILODCCATMW2OHGXSA4G","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"The Calibration Floor: Format Repair Can Masquerade as Self-Correction at Small-to-Mid Scale","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bo Qu, Licheng Wang, Mingguang Chen","submitted_at":"2026-08-05T01:55:33Z","abstract_excerpt":"Accuracy changes after language-model self-revision are usually interpreted as changes in reasoning. We show this can fail at the answer-extraction boundary, and test the failure causally rather than only observationally. Across Qwen3.5 (0.8B-9B), Gemma-4-12B, and two frontier models via API (Tencent Hy3, Nvidia Nemotron-3-Ultra-550B) in 29 primary cells plus a frontier arm, we decompose the always-revise accuracy shift into a content margin (both answers parseable) and format-recovery/loss margins (parseability changes). On 12 cells with meaningful unparseable-answer rates, format effects exc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04355","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/2608.04355/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":"2026-08-06T01:33:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Qf2YRzAaDAqn7hjE4a/SU/I+2WP6DsyIFfwUVYbuu1o3qg9H/NoOfl0pzB8zxfuCXBqotxzKTmrZO5hBUeAkCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T02:21:02.306196Z"},"content_sha256":"486764ff2b2f27f22b5f26baf5d21b002a6bbbe2f336b63822f08cd41fa8cecf","schema_version":"1.0","event_id":"sha256:486764ff2b2f27f22b5f26baf5d21b002a6bbbe2f336b63822f08cd41fa8cecf"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/BJH4GYKILODCCATMW2OHGXSA4G/bundle.json","state_url":"https://pith.science/pith/BJH4GYKILODCCATMW2OHGXSA4G/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/BJH4GYKILODCCATMW2OHGXSA4G/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":"2026-08-09T02:21:02Z","links":{"resolver":"https://pith.science/pith/BJH4GYKILODCCATMW2OHGXSA4G","bundle":"https://pith.science/pith/BJH4GYKILODCCATMW2OHGXSA4G/bundle.json","state":"https://pith.science/pith/BJH4GYKILODCCATMW2OHGXSA4G/state.json","well_known_bundle":"https://pith.science/.well-known/pith/BJH4GYKILODCCATMW2OHGXSA4G/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:BJH4GYKILODCCATMW2OHGXSA4G","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":"bcd59d1670722a85d00516e6fff82551de99cd310d63d47bdbdfccd4b7cdd050","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T01:55:33Z","title_canon_sha256":"67bb7e2dc1a1c9b40f1894b1c3a6294f2c25443dfa36c1a7587a0513aa38a115"},"schema_version":"1.0","source":{"id":"2608.04355","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.04355","created_at":"2026-08-06T01:33:35Z"},{"alias_kind":"arxiv_version","alias_value":"2608.04355v1","created_at":"2026-08-06T01:33:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04355","created_at":"2026-08-06T01:33:35Z"},{"alias_kind":"pith_short_12","alias_value":"BJH4GYKILODC","created_at":"2026-08-06T01:33:35Z"},{"alias_kind":"pith_short_16","alias_value":"BJH4GYKILODCCATM","created_at":"2026-08-06T01:33:35Z"},{"alias_kind":"pith_short_8","alias_value":"BJH4GYKI","created_at":"2026-08-06T01:33:35Z"}],"graph_snapshots":[{"event_id":"sha256:486764ff2b2f27f22b5f26baf5d21b002a6bbbe2f336b63822f08cd41fa8cecf","target":"graph","created_at":"2026-08-06T01:33:35Z","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/2608.04355/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Accuracy changes after language-model self-revision are usually interpreted as changes in reasoning. We show this can fail at the answer-extraction boundary, and test the failure causally rather than only observationally. Across Qwen3.5 (0.8B-9B), Gemma-4-12B, and two frontier models via API (Tencent Hy3, Nvidia Nemotron-3-Ultra-550B) in 29 primary cells plus a frontier arm, we decompose the always-revise accuracy shift into a content margin (both answers parseable) and format-recovery/loss margins (parseability changes). On 12 cells with meaningful unparseable-answer rates, format effects exc","authors_text":"Bo Qu, Licheng Wang, Mingguang Chen","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T01:55:33Z","title":"The Calibration Floor: Format Repair Can Masquerade as Self-Correction at Small-to-Mid Scale"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04355","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":"sha256:711fee77ed32c8334272391086b58298dc475f0197c180ae81cc4a1bf830c924","target":"record","created_at":"2026-08-06T01:33:35Z","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":"bcd59d1670722a85d00516e6fff82551de99cd310d63d47bdbdfccd4b7cdd050","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T01:55:33Z","title_canon_sha256":"67bb7e2dc1a1c9b40f1894b1c3a6294f2c25443dfa36c1a7587a0513aa38a115"},"schema_version":"1.0","source":{"id":"2608.04355","kind":"arxiv","version":1}},"canonical_sha256":"0a4fc361485b8621026cb69c735e40e1a96dd5e347f010fdc0322e4cdce11017","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0a4fc361485b8621026cb69c735e40e1a96dd5e347f010fdc0322e4cdce11017","first_computed_at":"2026-08-06T01:33:35.047211Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-06T01:33:35.047211Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"kIzHrTZCePGmSnalIHD95uK1o42V7oh/bkB6As+xGyGFTZFhJEhLa7cqLrAog3xpS30qYUNNz81y5Nd+xHZdCw==","signature_status":"signed_v1","signed_at":"2026-08-06T01:33:35.048700Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.04355","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:711fee77ed32c8334272391086b58298dc475f0197c180ae81cc4a1bf830c924","sha256:486764ff2b2f27f22b5f26baf5d21b002a6bbbe2f336b63822f08cd41fa8cecf"],"state_sha256":"1dc1d988e578228332acb2e8d1a59a48737e53dada3d1c85135086a952589eaa"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4RDsba52f4YN6dn4wsL0EqklyQPsWSSQpPnSx7cXtIQdialwNPCXFodxEvfFi5vC98T+BvpGWoses1hcHqDzCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T02:21:02.310935Z","bundle_sha256":"12ae0e57b78791e99453fa6b539f8b3f4230f5078c47d2d06040d4e4651c05e9"}}