{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:I2CZGMKF6L24ITLS2V722MW67Z","short_pith_number":"pith:I2CZGMKF","schema_version":"1.0","canonical_sha256":"4685933145f2f5c44d72d57fad32defe4d3db669f1d1f7593d00102d0607a6b9","source":{"kind":"arxiv","id":"2505.15585","version":1},"attestation_state":"computed","paper":{"title":"MIRB: Mathematical Information Retrieval Benchmark","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.IR","authors_text":"Bin Dong, Haocheng Ju","submitted_at":"2025-05-21T14:40:27Z","abstract_excerpt":"Mathematical Information Retrieval (MIR) is the task of retrieving information from mathematical documents and plays a key role in various applications, including theorem search in mathematical libraries, answer retrieval on math forums, and premise selection in automated theorem proving. However, a unified benchmark for evaluating these diverse retrieval tasks has been lacking. In this paper, we introduce MIRB (Mathematical Information Retrieval Benchmark) to assess the MIR capabilities of retrieval models. MIRB includes four tasks: semantic statement retrieval, question-answer retrieval, pre"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2505.15585","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2025-05-21T14:40:27Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"5574569e0a854b7bdb18279d7935ff2669d29b86c5edbfc9a526065753816ffd","abstract_canon_sha256":"4622297ff4f91e6629cbd41f2d8269dedcd817f0de92257f4b771ee120641cbf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:06:48.217901Z","signature_b64":"buCuWQ9tJ1vWwtOv8LJtC4ccY+7hQTBuPdis3ZR674eS0cGNXUIeYEh4VeMC7zdMloRx9earnqWUvE1Bcem6Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4685933145f2f5c44d72d57fad32defe4d3db669f1d1f7593d00102d0607a6b9","last_reissued_at":"2026-07-05T11:06:48.217426Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:06:48.217426Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MIRB: Mathematical Information Retrieval Benchmark","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.IR","authors_text":"Bin Dong, Haocheng Ju","submitted_at":"2025-05-21T14:40:27Z","abstract_excerpt":"Mathematical Information Retrieval (MIR) is the task of retrieving information from mathematical documents and plays a key role in various applications, including theorem search in mathematical libraries, answer retrieval on math forums, and premise selection in automated theorem proving. However, a unified benchmark for evaluating these diverse retrieval tasks has been lacking. In this paper, we introduce MIRB (Mathematical Information Retrieval Benchmark) to assess the MIR capabilities of retrieval models. MIRB includes four tasks: semantic statement retrieval, question-answer retrieval, pre"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.15585","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/2505.15585/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2505.15585","created_at":"2026-07-05T11:06:48.217487+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.15585v1","created_at":"2026-07-05T11:06:48.217487+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.15585","created_at":"2026-07-05T11:06:48.217487+00:00"},{"alias_kind":"pith_short_12","alias_value":"I2CZGMKF6L24","created_at":"2026-07-05T11:06:48.217487+00:00"},{"alias_kind":"pith_short_16","alias_value":"I2CZGMKF6L24ITLS","created_at":"2026-07-05T11:06:48.217487+00:00"},{"alias_kind":"pith_short_8","alias_value":"I2CZGMKF","created_at":"2026-07-05T11:06:48.217487+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.13137","citing_title":"LeanSearch v2: Global Premise Retrieval for Lean 4 Theorem Proving","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13137","citing_title":"LeanSearch v2: Global Premise Retrieval for Lean 4 Theorem Proving","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/I2CZGMKF6L24ITLS2V722MW67Z","json":"https://pith.science/pith/I2CZGMKF6L24ITLS2V722MW67Z.json","graph_json":"https://pith.science/api/pith-number/I2CZGMKF6L24ITLS2V722MW67Z/graph.json","events_json":"https://pith.science/api/pith-number/I2CZGMKF6L24ITLS2V722MW67Z/events.json","paper":"https://pith.science/paper/I2CZGMKF"},"agent_actions":{"view_html":"https://pith.science/pith/I2CZGMKF6L24ITLS2V722MW67Z","download_json":"https://pith.science/pith/I2CZGMKF6L24ITLS2V722MW67Z.json","view_paper":"https://pith.science/paper/I2CZGMKF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.15585&json=true","fetch_graph":"https://pith.science/api/pith-number/I2CZGMKF6L24ITLS2V722MW67Z/graph.json","fetch_events":"https://pith.science/api/pith-number/I2CZGMKF6L24ITLS2V722MW67Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/I2CZGMKF6L24ITLS2V722MW67Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/I2CZGMKF6L24ITLS2V722MW67Z/action/storage_attestation","attest_author":"https://pith.science/pith/I2CZGMKF6L24ITLS2V722MW67Z/action/author_attestation","sign_citation":"https://pith.science/pith/I2CZGMKF6L24ITLS2V722MW67Z/action/citation_signature","submit_replication":"https://pith.science/pith/I2CZGMKF6L24ITLS2V722MW67Z/action/replication_record"}},"created_at":"2026-07-05T11:06:48.217487+00:00","updated_at":"2026-07-05T11:06:48.217487+00:00"}