{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:OBLJTUBJ6WDZIJ55HBVFK6LI3Q","short_pith_number":"pith:OBLJTUBJ","schema_version":"1.0","canonical_sha256":"705699d029f5879427bd386a557968dc02c2a8ccb14a881fcc35eadca34a1f03","source":{"kind":"arxiv","id":"2506.04535","version":1},"attestation_state":"computed","paper":{"title":"BSBench: will your LLM find the largest prime number?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"K. O. T. Erziev","submitted_at":"2025-06-05T00:59:16Z","abstract_excerpt":"We propose that benchmarking LLMs on questions which have no reasonable answer actually isn't as silly as it sounds. We also present a benchmark that allows such testing and a method to modify the existing datasets, and discover that existing models demonstrate a performance far from the perfect on such questions. Our code and data artifacts are available at https://github.com/L3G5/impossible-bench"},"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":"2506.04535","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-05T00:59:16Z","cross_cats_sorted":[],"title_canon_sha256":"85645a7da71b23847084b55b8b40b33e6a451879a71dce5f4817f7b397da4132","abstract_canon_sha256":"06fb92bde0d65ace5fcd0af29c50d21086dedc4a7aee7adc5e89fcf6bc1c224a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:16:16.228129Z","signature_b64":"JlOormZ/fRKvqPCH6cfjaY8lImFZCdW+FldWptkZ1zFoDCUlmHBBeiNHhKcFaAmhy/4gSxue/CJrNJlhvPNsBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"705699d029f5879427bd386a557968dc02c2a8ccb14a881fcc35eadca34a1f03","last_reissued_at":"2026-07-05T11:16:16.227762Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:16:16.227762Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BSBench: will your LLM find the largest prime number?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"K. O. T. Erziev","submitted_at":"2025-06-05T00:59:16Z","abstract_excerpt":"We propose that benchmarking LLMs on questions which have no reasonable answer actually isn't as silly as it sounds. We also present a benchmark that allows such testing and a method to modify the existing datasets, and discover that existing models demonstrate a performance far from the perfect on such questions. Our code and data artifacts are available at https://github.com/L3G5/impossible-bench"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.04535","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/2506.04535/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":"2506.04535","created_at":"2026-07-05T11:16:16.227819+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.04535v1","created_at":"2026-07-05T11:16:16.227819+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.04535","created_at":"2026-07-05T11:16:16.227819+00:00"},{"alias_kind":"pith_short_12","alias_value":"OBLJTUBJ6WDZ","created_at":"2026-07-05T11:16:16.227819+00:00"},{"alias_kind":"pith_short_16","alias_value":"OBLJTUBJ6WDZIJ55","created_at":"2026-07-05T11:16:16.227819+00:00"},{"alias_kind":"pith_short_8","alias_value":"OBLJTUBJ","created_at":"2026-07-05T11:16:16.227819+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OBLJTUBJ6WDZIJ55HBVFK6LI3Q","json":"https://pith.science/pith/OBLJTUBJ6WDZIJ55HBVFK6LI3Q.json","graph_json":"https://pith.science/api/pith-number/OBLJTUBJ6WDZIJ55HBVFK6LI3Q/graph.json","events_json":"https://pith.science/api/pith-number/OBLJTUBJ6WDZIJ55HBVFK6LI3Q/events.json","paper":"https://pith.science/paper/OBLJTUBJ"},"agent_actions":{"view_html":"https://pith.science/pith/OBLJTUBJ6WDZIJ55HBVFK6LI3Q","download_json":"https://pith.science/pith/OBLJTUBJ6WDZIJ55HBVFK6LI3Q.json","view_paper":"https://pith.science/paper/OBLJTUBJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.04535&json=true","fetch_graph":"https://pith.science/api/pith-number/OBLJTUBJ6WDZIJ55HBVFK6LI3Q/graph.json","fetch_events":"https://pith.science/api/pith-number/OBLJTUBJ6WDZIJ55HBVFK6LI3Q/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OBLJTUBJ6WDZIJ55HBVFK6LI3Q/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OBLJTUBJ6WDZIJ55HBVFK6LI3Q/action/storage_attestation","attest_author":"https://pith.science/pith/OBLJTUBJ6WDZIJ55HBVFK6LI3Q/action/author_attestation","sign_citation":"https://pith.science/pith/OBLJTUBJ6WDZIJ55HBVFK6LI3Q/action/citation_signature","submit_replication":"https://pith.science/pith/OBLJTUBJ6WDZIJ55HBVFK6LI3Q/action/replication_record"}},"created_at":"2026-07-05T11:16:16.227819+00:00","updated_at":"2026-07-05T11:16:16.227819+00:00"}