{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:UMCJKFGFX6CMTICOFJV4CVI7UZ","short_pith_number":"pith:UMCJKFGF","canonical_record":{"source":{"id":"2508.11252","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-08-15T06:42:00Z","cross_cats_sorted":["cs.CL","cs.IR"],"title_canon_sha256":"c06d1df8e075f23e0b8498cd56b62f2269c0c39ab48e25d17418cf72f8bb374d","abstract_canon_sha256":"5eb11e07f25bd0748db537c933a93ff58b2e79a32e56ce1ef1e168a30bf47d89"},"schema_version":"1.0"},"canonical_sha256":"a3049514c5bf84c9a04e2a6bc1551fa66c01f6ae10c036f541ad4798e27c6d7c","source":{"kind":"arxiv","id":"2508.11252","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.11252","created_at":"2026-07-05T11:54:21Z"},{"alias_kind":"arxiv_version","alias_value":"2508.11252v1","created_at":"2026-07-05T11:54:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.11252","created_at":"2026-07-05T11:54:21Z"},{"alias_kind":"pith_short_12","alias_value":"UMCJKFGFX6CM","created_at":"2026-07-05T11:54:21Z"},{"alias_kind":"pith_short_16","alias_value":"UMCJKFGFX6CMTICO","created_at":"2026-07-05T11:54:21Z"},{"alias_kind":"pith_short_8","alias_value":"UMCJKFGF","created_at":"2026-07-05T11:54:21Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:UMCJKFGFX6CMTICOFJV4CVI7UZ","target":"record","payload":{"canonical_record":{"source":{"id":"2508.11252","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-08-15T06:42:00Z","cross_cats_sorted":["cs.CL","cs.IR"],"title_canon_sha256":"c06d1df8e075f23e0b8498cd56b62f2269c0c39ab48e25d17418cf72f8bb374d","abstract_canon_sha256":"5eb11e07f25bd0748db537c933a93ff58b2e79a32e56ce1ef1e168a30bf47d89"},"schema_version":"1.0"},"canonical_sha256":"a3049514c5bf84c9a04e2a6bc1551fa66c01f6ae10c036f541ad4798e27c6d7c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:54:21.402900Z","signature_b64":"fJnFgTgxUn83uQwqH1NUEcEoFBr7zWV7XhccoahwA7vQjMT58T1u9yUdMqZtjLc2ic+IDWX/K7uwa/fY4YjVCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a3049514c5bf84c9a04e2a6bc1551fa66c01f6ae10c036f541ad4798e27c6d7c","last_reissued_at":"2026-07-05T11:54:21.402483Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:54:21.402483Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2508.11252","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:54:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"icW77FkujVHcsjQ2eYu6SZOyeyc0xeiKvvSL02IzSNVY/8DvCG0TVQnrh4DmEh5FJ99vNGzRjYqn1cqD0VEcAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T03:40:16.814654Z"},"content_sha256":"5b46f85b13892368cbaaadb3a2dce1bf249e13d153d19d72df93bc80a46484ac","schema_version":"1.0","event_id":"sha256:5b46f85b13892368cbaaadb3a2dce1bf249e13d153d19d72df93bc80a46484ac"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:UMCJKFGFX6CMTICOFJV4CVI7UZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Beyond Solving Math Quiz: Evaluating the Ability of Large Reasoning Models to Ask for Information","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.IR"],"primary_cat":"cs.AI","authors_text":"Bowen Qin, Chen Huang, Duanyu Feng, Wenqiang Lei, Xi Yang, Youcheng Huang","submitted_at":"2025-08-15T06:42:00Z","abstract_excerpt":"Large Reasoning Models (LRMs) have demonstrated remarkable problem-solving abilities in mathematics, as evaluated by existing benchmarks exclusively on well-defined problems. However, such evaluation setup constitutes a critical gap, since a genuine intelligent agent should not only solve problems (as a math quiz solver), but also be able~to ask for information when the problems lack sufficient information, enabling proactivity in responding users' requests. To bridge such gap, we proposes a new dataset consisting of two types of incomplete problems with diverse contexts. Based on the dataset,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.11252","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/2508.11252/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:54:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NmmZBGV51Z0CEvvcQLBIb4XGV7Xn4nCwElTASJIDZ56vVCtmNTX5HdWfKijNFO4YlT2PhPi2Bz2874GN4lW9Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T03:40:16.815357Z"},"content_sha256":"864c11481442c4db9d20e6a8b0f9b71bd60f3666a8e845a5ebc10c4749d7064e","schema_version":"1.0","event_id":"sha256:864c11481442c4db9d20e6a8b0f9b71bd60f3666a8e845a5ebc10c4749d7064e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/UMCJKFGFX6CMTICOFJV4CVI7UZ/bundle.json","state_url":"https://pith.science/pith/UMCJKFGFX6CMTICOFJV4CVI7UZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/UMCJKFGFX6CMTICOFJV4CVI7UZ/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-07T03:40:16Z","links":{"resolver":"https://pith.science/pith/UMCJKFGFX6CMTICOFJV4CVI7UZ","bundle":"https://pith.science/pith/UMCJKFGFX6CMTICOFJV4CVI7UZ/bundle.json","state":"https://pith.science/pith/UMCJKFGFX6CMTICOFJV4CVI7UZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/UMCJKFGFX6CMTICOFJV4CVI7UZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:UMCJKFGFX6CMTICOFJV4CVI7UZ","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":"5eb11e07f25bd0748db537c933a93ff58b2e79a32e56ce1ef1e168a30bf47d89","cross_cats_sorted":["cs.CL","cs.IR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-08-15T06:42:00Z","title_canon_sha256":"c06d1df8e075f23e0b8498cd56b62f2269c0c39ab48e25d17418cf72f8bb374d"},"schema_version":"1.0","source":{"id":"2508.11252","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.11252","created_at":"2026-07-05T11:54:21Z"},{"alias_kind":"arxiv_version","alias_value":"2508.11252v1","created_at":"2026-07-05T11:54:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.11252","created_at":"2026-07-05T11:54:21Z"},{"alias_kind":"pith_short_12","alias_value":"UMCJKFGFX6CM","created_at":"2026-07-05T11:54:21Z"},{"alias_kind":"pith_short_16","alias_value":"UMCJKFGFX6CMTICO","created_at":"2026-07-05T11:54:21Z"},{"alias_kind":"pith_short_8","alias_value":"UMCJKFGF","created_at":"2026-07-05T11:54:21Z"}],"graph_snapshots":[{"event_id":"sha256:864c11481442c4db9d20e6a8b0f9b71bd60f3666a8e845a5ebc10c4749d7064e","target":"graph","created_at":"2026-07-05T11:54:21Z","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/2508.11252/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Reasoning Models (LRMs) have demonstrated remarkable problem-solving abilities in mathematics, as evaluated by existing benchmarks exclusively on well-defined problems. However, such evaluation setup constitutes a critical gap, since a genuine intelligent agent should not only solve problems (as a math quiz solver), but also be able~to ask for information when the problems lack sufficient information, enabling proactivity in responding users' requests. To bridge such gap, we proposes a new dataset consisting of two types of incomplete problems with diverse contexts. Based on the dataset,","authors_text":"Bowen Qin, Chen Huang, Duanyu Feng, Wenqiang Lei, Xi Yang, Youcheng Huang","cross_cats":["cs.CL","cs.IR"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-08-15T06:42:00Z","title":"Beyond Solving Math Quiz: Evaluating the Ability of Large Reasoning Models to Ask for Information"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.11252","kind":"arxiv","version":1},"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":"sha256:5b46f85b13892368cbaaadb3a2dce1bf249e13d153d19d72df93bc80a46484ac","target":"record","created_at":"2026-07-05T11:54:21Z","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":"5eb11e07f25bd0748db537c933a93ff58b2e79a32e56ce1ef1e168a30bf47d89","cross_cats_sorted":["cs.CL","cs.IR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-08-15T06:42:00Z","title_canon_sha256":"c06d1df8e075f23e0b8498cd56b62f2269c0c39ab48e25d17418cf72f8bb374d"},"schema_version":"1.0","source":{"id":"2508.11252","kind":"arxiv","version":1}},"canonical_sha256":"a3049514c5bf84c9a04e2a6bc1551fa66c01f6ae10c036f541ad4798e27c6d7c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a3049514c5bf84c9a04e2a6bc1551fa66c01f6ae10c036f541ad4798e27c6d7c","first_computed_at":"2026-07-05T11:54:21.402483Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:54:21.402483Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"fJnFgTgxUn83uQwqH1NUEcEoFBr7zWV7XhccoahwA7vQjMT58T1u9yUdMqZtjLc2ic+IDWX/K7uwa/fY4YjVCw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:54:21.402900Z","signed_message":"canonical_sha256_bytes"},"source_id":"2508.11252","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5b46f85b13892368cbaaadb3a2dce1bf249e13d153d19d72df93bc80a46484ac","sha256:864c11481442c4db9d20e6a8b0f9b71bd60f3666a8e845a5ebc10c4749d7064e"],"state_sha256":"374b86966686e0b399090862cd05819bca3599807381753113b46ae98c86591a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"neMjddzbAsFOYxP8D6EdXp4Nb319IVKeoswwG1/0UQpFSZiWHxfO1uhdvRkYWEI54th8SxwMMWb10l0mOSoQBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T03:40:16.820475Z","bundle_sha256":"c6b5553bac4926d1b47a7f24a4294348e9d5e4d7c2f5d4fc2c7747fdae5483fd"}}