{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:MMUZEXYLMKPEXX3655UHPTYM5C","short_pith_number":"pith:MMUZEXYL","schema_version":"1.0","canonical_sha256":"6329925f0b629e4bdf7eef6877cf0ce89ac1daa949f18b3d34f60eca81efbaff","source":{"kind":"arxiv","id":"2502.03358","version":2},"attestation_state":"computed","paper":{"title":"Minerva: A Programmable Memory Test Benchmark for Language Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Menglin Xia, Reza Shokri, Saravan Rajmohan, Victor Ruehle","submitted_at":"2025-02-05T16:53:45Z","abstract_excerpt":"How effectively can LLM-based AI assistants utilize their memory (context) to perform various tasks? Traditional data benchmarks, which are often manually crafted, suffer from several limitations: they are static, susceptible to overfitting, difficult to interpret, and lack actionable insights--failing to pinpoint the specific capabilities a model lacks when it does not pass a test. In this paper, we present a framework for automatically generating a comprehensive set of tests to evaluate models' abilities to use their memory effectively. Our framework extends the range of capability tests bey"},"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":"2502.03358","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-05T16:53:45Z","cross_cats_sorted":[],"title_canon_sha256":"6910500003e7fb7262fe42b048242c5231358424e5dace89d816d637d513eba5","abstract_canon_sha256":"bf76767bb4402e756114ab2ac074c0692659424124243a64102f6336d2557f31"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:17.549188Z","signature_b64":"N3lv1sw2gRIwXEk6gsKC0vNy6al9qzrFOUITv5x87v1Gw9j2uiOMXgVDuT5zGvzGOj9u6AtrePjYggpvFfn7CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6329925f0b629e4bdf7eef6877cf0ce89ac1daa949f18b3d34f60eca81efbaff","last_reissued_at":"2026-07-05T11:18:17.548676Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:17.548676Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Minerva: A Programmable Memory Test Benchmark for Language Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Menglin Xia, Reza Shokri, Saravan Rajmohan, Victor Ruehle","submitted_at":"2025-02-05T16:53:45Z","abstract_excerpt":"How effectively can LLM-based AI assistants utilize their memory (context) to perform various tasks? Traditional data benchmarks, which are often manually crafted, suffer from several limitations: they are static, susceptible to overfitting, difficult to interpret, and lack actionable insights--failing to pinpoint the specific capabilities a model lacks when it does not pass a test. In this paper, we present a framework for automatically generating a comprehensive set of tests to evaluate models' abilities to use their memory effectively. Our framework extends the range of capability tests bey"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.03358","kind":"arxiv","version":2},"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/2502.03358/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":"2502.03358","created_at":"2026-07-05T11:18:17.548743+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.03358v2","created_at":"2026-07-05T11:18:17.548743+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.03358","created_at":"2026-07-05T11:18:17.548743+00:00"},{"alias_kind":"pith_short_12","alias_value":"MMUZEXYLMKPE","created_at":"2026-07-05T11:18:17.548743+00:00"},{"alias_kind":"pith_short_16","alias_value":"MMUZEXYLMKPEXX36","created_at":"2026-07-05T11:18:17.548743+00:00"},{"alias_kind":"pith_short_8","alias_value":"MMUZEXYL","created_at":"2026-07-05T11:18:17.548743+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.18182","citing_title":"SCOPE: Stochastic and Counterbiased Option Placement for Evaluating Large Language Models","ref_index":68,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MMUZEXYLMKPEXX3655UHPTYM5C","json":"https://pith.science/pith/MMUZEXYLMKPEXX3655UHPTYM5C.json","graph_json":"https://pith.science/api/pith-number/MMUZEXYLMKPEXX3655UHPTYM5C/graph.json","events_json":"https://pith.science/api/pith-number/MMUZEXYLMKPEXX3655UHPTYM5C/events.json","paper":"https://pith.science/paper/MMUZEXYL"},"agent_actions":{"view_html":"https://pith.science/pith/MMUZEXYLMKPEXX3655UHPTYM5C","download_json":"https://pith.science/pith/MMUZEXYLMKPEXX3655UHPTYM5C.json","view_paper":"https://pith.science/paper/MMUZEXYL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.03358&json=true","fetch_graph":"https://pith.science/api/pith-number/MMUZEXYLMKPEXX3655UHPTYM5C/graph.json","fetch_events":"https://pith.science/api/pith-number/MMUZEXYLMKPEXX3655UHPTYM5C/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MMUZEXYLMKPEXX3655UHPTYM5C/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MMUZEXYLMKPEXX3655UHPTYM5C/action/storage_attestation","attest_author":"https://pith.science/pith/MMUZEXYLMKPEXX3655UHPTYM5C/action/author_attestation","sign_citation":"https://pith.science/pith/MMUZEXYLMKPEXX3655UHPTYM5C/action/citation_signature","submit_replication":"https://pith.science/pith/MMUZEXYLMKPEXX3655UHPTYM5C/action/replication_record"}},"created_at":"2026-07-05T11:18:17.548743+00:00","updated_at":"2026-07-05T11:18:17.548743+00:00"}