{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:6WXWBLLSDPC6RB5OOS263GCSYJ","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":"fcc76777a217476db059cba368e2818a9a4878d890f4202b6232051e1e62ead3","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.PL","submitted_at":"2025-07-16T03:11:43Z","title_canon_sha256":"479b6938cf3149b6e2d3c81ab6f2f24a52965222582ee610dc47af85f3ec49a4"},"schema_version":"1.0","source":{"id":"2507.22069","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.22069","created_at":"2026-07-05T11:46:00Z"},{"alias_kind":"arxiv_version","alias_value":"2507.22069v2","created_at":"2026-07-05T11:46:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.22069","created_at":"2026-07-05T11:46:00Z"},{"alias_kind":"pith_short_12","alias_value":"6WXWBLLSDPC6","created_at":"2026-07-05T11:46:00Z"},{"alias_kind":"pith_short_16","alias_value":"6WXWBLLSDPC6RB5O","created_at":"2026-07-05T11:46:00Z"},{"alias_kind":"pith_short_8","alias_value":"6WXWBLLS","created_at":"2026-07-05T11:46:00Z"}],"graph_snapshots":[{"event_id":"sha256:e25178a2a754ba6b43857712f4fe8f56373d68aa51d9b59e4138e606234c395d","target":"graph","created_at":"2026-07-05T11:46:00Z","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/2507.22069/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reusing established theorems and formulas is central to mathematical problem solving, serving as essential building blocks for tackling increasingly complex challenges. Recent work, TroVE, argues that code-generating Large Language Models (LLMs) can benefit similarly on the MATH benchmark by inducing and reusing higher-level toolboxes. By allocating computational budget across an ensemble of three modes -- directly generating code, creating tools, and reusing tools -- TroVE claims to outperform a PRIMITIVE baseline that only performs direct generation. However, recent analysis (Berlot-Attwell ","authors_text":"Christian Bartelt, Ian Berlot-Attwell, Janis Zenkner, Tobias Sesterhenn","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.PL","submitted_at":"2025-07-16T03:11:43Z","title":"A Compute-Matched Re-Evaluation of TroVE on MATH"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.22069","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:a55f96ed97b08feee4bd530c3c814b84bc45e1e46be0cdd810f276ee4d70be47","target":"record","created_at":"2026-07-05T11:46:00Z","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":"fcc76777a217476db059cba368e2818a9a4878d890f4202b6232051e1e62ead3","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.PL","submitted_at":"2025-07-16T03:11:43Z","title_canon_sha256":"479b6938cf3149b6e2d3c81ab6f2f24a52965222582ee610dc47af85f3ec49a4"},"schema_version":"1.0","source":{"id":"2507.22069","kind":"arxiv","version":2}},"canonical_sha256":"f5af60ad721bc5e887ae74b5ed9852c26d70fa4972b71131fa8eb3fb2730cf62","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f5af60ad721bc5e887ae74b5ed9852c26d70fa4972b71131fa8eb3fb2730cf62","first_computed_at":"2026-07-05T11:46:00.384528Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:46:00.384528Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"RmqlWc9epijwM5YIb0G/83qzlKV5Yxb+uZw1CBkTcPX6f3dXmFXWrzvU4Qd4bgtqiBNONBfbtklOyrkZi85jAA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:46:00.385055Z","signed_message":"canonical_sha256_bytes"},"source_id":"2507.22069","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a55f96ed97b08feee4bd530c3c814b84bc45e1e46be0cdd810f276ee4d70be47","sha256:e25178a2a754ba6b43857712f4fe8f56373d68aa51d9b59e4138e606234c395d"],"state_sha256":"5046541f5f23f2ecf5fbb5cb5ff272dfc56cc15b06cd5f4c1d12a79c80dda374"}