{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:RBGTRZSTRJZVC3ETREUVCVYMB2","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":"14bb73a720ccaf0c5e93236b56916fcb23a79ccfc72609d4730a6d4895023a75","cross_cats_sorted":["cs.LG","cs.LO","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2020-07-06T17:55:33Z","title_canon_sha256":"49f5ac2651e0c8169333769fa89bc815029c13ed42ccb787180c26114af35a71"},"schema_version":"1.0","source":{"id":"2007.02924","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2007.02924","created_at":"2026-07-05T02:28:50Z"},{"alias_kind":"arxiv_version","alias_value":"2007.02924v2","created_at":"2026-07-05T02:28:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2007.02924","created_at":"2026-07-05T02:28:50Z"},{"alias_kind":"pith_short_12","alias_value":"RBGTRZSTRJZV","created_at":"2026-07-05T02:28:50Z"},{"alias_kind":"pith_short_16","alias_value":"RBGTRZSTRJZVC3ET","created_at":"2026-07-05T02:28:50Z"},{"alias_kind":"pith_short_8","alias_value":"RBGTRZST","created_at":"2026-07-05T02:28:50Z"}],"graph_snapshots":[{"event_id":"sha256:8f5616b9515c6d121a9945ce381b274dbd17330e600975eb00a95a2031c7cb14","target":"graph","created_at":"2026-07-05T02:28:50Z","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/2007.02924/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In learning-assisted theorem proving, one of the most critical challenges is to generalize to theorems unlike those seen at training time. In this paper, we introduce INT, an INequality Theorem proving benchmark, specifically designed to test agents' generalization ability. INT is based on a procedure for generating theorems and proofs; this procedure's knobs allow us to measure 6 different types of generalization, each reflecting a distinct challenge characteristic to automated theorem proving. In addition, unlike prior benchmarks for learning-assisted theorem proving, INT provides a lightwei","authors_text":"Albert Qiaochu Jiang, Jimmy Ba, Roger Grosse, Yuhuai Wu","cross_cats":["cs.LG","cs.LO","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2020-07-06T17:55:33Z","title":"INT: An Inequality Benchmark for Evaluating Generalization in Theorem Proving"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2007.02924","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":"sha256:4a0a7f529042259108e0c489773df827c6f7a1614494d259e742342b5e11a275","target":"record","created_at":"2026-07-05T02:28:50Z","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":"14bb73a720ccaf0c5e93236b56916fcb23a79ccfc72609d4730a6d4895023a75","cross_cats_sorted":["cs.LG","cs.LO","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2020-07-06T17:55:33Z","title_canon_sha256":"49f5ac2651e0c8169333769fa89bc815029c13ed42ccb787180c26114af35a71"},"schema_version":"1.0","source":{"id":"2007.02924","kind":"arxiv","version":2}},"canonical_sha256":"884d38e6538a73516c93892951570c0e8af953a25c01079883d947065d8f8033","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"884d38e6538a73516c93892951570c0e8af953a25c01079883d947065d8f8033","first_computed_at":"2026-07-05T02:28:50.182200Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T02:28:50.182200Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"UOugto5lSSIxU+96hXu4LQLDfXhaLUUfkUitQ8FDMkp5DuIrxOAYAK3s+NZ+dJHYFCPZz0dGFofuW6FyZX/mDA==","signature_status":"signed_v1","signed_at":"2026-07-05T02:28:50.182631Z","signed_message":"canonical_sha256_bytes"},"source_id":"2007.02924","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4a0a7f529042259108e0c489773df827c6f7a1614494d259e742342b5e11a275","sha256:8f5616b9515c6d121a9945ce381b274dbd17330e600975eb00a95a2031c7cb14"],"state_sha256":"a43bddd08504874be12419e9b6eed2f1f7f683087b709340c35656773feb61c9"}