{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:L5YDC4IWIAPMPNITZZ6RIXUAYD","short_pith_number":"pith:L5YDC4IW","canonical_record":{"source":{"id":"2406.12641","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-18T14:08:01Z","cross_cats_sorted":[],"title_canon_sha256":"ce5f6b17e77ac03f3ffa4bfc25ea245052eaf39fcaa2e7b133bcd01b45ccf4d6","abstract_canon_sha256":"c0d5b67eef1dc0c5a0aa8139515b0530a8917a46b40e48b5883510ac7e07c362"},"schema_version":"1.0"},"canonical_sha256":"5f70317116401ec7b513ce7d145e80c0c97664d66e5878134100b7eb4f3f99d4","source":{"kind":"arxiv","id":"2406.12641","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.12641","created_at":"2026-07-05T09:33:33Z"},{"alias_kind":"arxiv_version","alias_value":"2406.12641v2","created_at":"2026-07-05T09:33:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.12641","created_at":"2026-07-05T09:33:33Z"},{"alias_kind":"pith_short_12","alias_value":"L5YDC4IWIAPM","created_at":"2026-07-05T09:33:33Z"},{"alias_kind":"pith_short_16","alias_value":"L5YDC4IWIAPMPNIT","created_at":"2026-07-05T09:33:33Z"},{"alias_kind":"pith_short_8","alias_value":"L5YDC4IW","created_at":"2026-07-05T09:33:33Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:L5YDC4IWIAPMPNITZZ6RIXUAYD","target":"record","payload":{"canonical_record":{"source":{"id":"2406.12641","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-18T14:08:01Z","cross_cats_sorted":[],"title_canon_sha256":"ce5f6b17e77ac03f3ffa4bfc25ea245052eaf39fcaa2e7b133bcd01b45ccf4d6","abstract_canon_sha256":"c0d5b67eef1dc0c5a0aa8139515b0530a8917a46b40e48b5883510ac7e07c362"},"schema_version":"1.0"},"canonical_sha256":"5f70317116401ec7b513ce7d145e80c0c97664d66e5878134100b7eb4f3f99d4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:33:33.786070Z","signature_b64":"6MpGbuBu54jeDgYreDSi4MqcdmwW+55eSU0Fiws9oVYsR3eoNTVGqs05ZhZBtU92apwBrs0KB3SrJgyZLEKTCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5f70317116401ec7b513ce7d145e80c0c97664d66e5878134100b7eb4f3f99d4","last_reissued_at":"2026-07-05T09:33:33.785578Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:33:33.785578Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.12641","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:33:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"heVJn3UJRsZiEOYIpt0QZQzwryVzNWWVvHdblQps81tGZU6BE8jFh+8qnLVDjUUXBD+AJ4xligrw5KIzH/4jAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T11:02:43.237941Z"},"content_sha256":"f3a2e04036e6a0ace040918e042e06253226dc29012f88d136f986c46d670023","schema_version":"1.0","event_id":"sha256:f3a2e04036e6a0ace040918e042e06253226dc29012f88d136f986c46d670023"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:L5YDC4IWIAPMPNITZZ6RIXUAYD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"DetectBench: Can Large Language Model Detect and Piece Together Implicit Evidence?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hongwei Feng, Jiangjie Chen, Lin Zhang, Shusen Wang, Wenhao Huang, Xiaoxuan Zhu, Yan Gao, Yanghua Xiao, Yikai Zhang, Zheyu Ye, Zhouhong Gu","submitted_at":"2024-06-18T14:08:01Z","abstract_excerpt":"Detecting evidence within the context is a key step in the process of reasoning task. Evaluating and enhancing the capabilities of LLMs in evidence detection will strengthen context-based reasoning performance. This paper proposes a benchmark called DetectBench for verifying the ability to detect and piece together implicit evidence within a long context. DetectBench contains 3,928 multiple-choice questions, with an average of 994 tokens per question. Each question contains an average of 4.55 pieces of implicit evidence, and solving the problem typically requires 7.62 logical jumps to find the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.12641","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/2406.12641/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":"2026-07-05T09:33:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WXlhsT/Ud0E6zkP2J6e+yhsnwQrC1wMU4erdpFjP3gvk2GaFz2wvOyBtJdK/1NIW8e48nVFSFwcQS/t4qTRKDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T11:02:43.238299Z"},"content_sha256":"3f60e4a973fdfd5569be0ed741298750fe47af2389ca0b047d180fb5e0bebc52","schema_version":"1.0","event_id":"sha256:3f60e4a973fdfd5569be0ed741298750fe47af2389ca0b047d180fb5e0bebc52"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/L5YDC4IWIAPMPNITZZ6RIXUAYD/bundle.json","state_url":"https://pith.science/pith/L5YDC4IWIAPMPNITZZ6RIXUAYD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/L5YDC4IWIAPMPNITZZ6RIXUAYD/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":"2026-07-07T11:02:43Z","links":{"resolver":"https://pith.science/pith/L5YDC4IWIAPMPNITZZ6RIXUAYD","bundle":"https://pith.science/pith/L5YDC4IWIAPMPNITZZ6RIXUAYD/bundle.json","state":"https://pith.science/pith/L5YDC4IWIAPMPNITZZ6RIXUAYD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/L5YDC4IWIAPMPNITZZ6RIXUAYD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:L5YDC4IWIAPMPNITZZ6RIXUAYD","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":"c0d5b67eef1dc0c5a0aa8139515b0530a8917a46b40e48b5883510ac7e07c362","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-18T14:08:01Z","title_canon_sha256":"ce5f6b17e77ac03f3ffa4bfc25ea245052eaf39fcaa2e7b133bcd01b45ccf4d6"},"schema_version":"1.0","source":{"id":"2406.12641","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.12641","created_at":"2026-07-05T09:33:33Z"},{"alias_kind":"arxiv_version","alias_value":"2406.12641v2","created_at":"2026-07-05T09:33:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.12641","created_at":"2026-07-05T09:33:33Z"},{"alias_kind":"pith_short_12","alias_value":"L5YDC4IWIAPM","created_at":"2026-07-05T09:33:33Z"},{"alias_kind":"pith_short_16","alias_value":"L5YDC4IWIAPMPNIT","created_at":"2026-07-05T09:33:33Z"},{"alias_kind":"pith_short_8","alias_value":"L5YDC4IW","created_at":"2026-07-05T09:33:33Z"}],"graph_snapshots":[{"event_id":"sha256:3f60e4a973fdfd5569be0ed741298750fe47af2389ca0b047d180fb5e0bebc52","target":"graph","created_at":"2026-07-05T09:33:33Z","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/2406.12641/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Detecting evidence within the context is a key step in the process of reasoning task. Evaluating and enhancing the capabilities of LLMs in evidence detection will strengthen context-based reasoning performance. This paper proposes a benchmark called DetectBench for verifying the ability to detect and piece together implicit evidence within a long context. DetectBench contains 3,928 multiple-choice questions, with an average of 994 tokens per question. Each question contains an average of 4.55 pieces of implicit evidence, and solving the problem typically requires 7.62 logical jumps to find the","authors_text":"Hongwei Feng, Jiangjie Chen, Lin Zhang, Shusen Wang, Wenhao Huang, Xiaoxuan Zhu, Yan Gao, Yanghua Xiao, Yikai Zhang, Zheyu Ye, Zhouhong Gu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-18T14:08:01Z","title":"DetectBench: Can Large Language Model Detect and Piece Together Implicit Evidence?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.12641","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:f3a2e04036e6a0ace040918e042e06253226dc29012f88d136f986c46d670023","target":"record","created_at":"2026-07-05T09:33:33Z","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":"c0d5b67eef1dc0c5a0aa8139515b0530a8917a46b40e48b5883510ac7e07c362","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-18T14:08:01Z","title_canon_sha256":"ce5f6b17e77ac03f3ffa4bfc25ea245052eaf39fcaa2e7b133bcd01b45ccf4d6"},"schema_version":"1.0","source":{"id":"2406.12641","kind":"arxiv","version":2}},"canonical_sha256":"5f70317116401ec7b513ce7d145e80c0c97664d66e5878134100b7eb4f3f99d4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5f70317116401ec7b513ce7d145e80c0c97664d66e5878134100b7eb4f3f99d4","first_computed_at":"2026-07-05T09:33:33.785578Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:33:33.785578Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"6MpGbuBu54jeDgYreDSi4MqcdmwW+55eSU0Fiws9oVYsR3eoNTVGqs05ZhZBtU92apwBrs0KB3SrJgyZLEKTCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:33:33.786070Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.12641","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f3a2e04036e6a0ace040918e042e06253226dc29012f88d136f986c46d670023","sha256:3f60e4a973fdfd5569be0ed741298750fe47af2389ca0b047d180fb5e0bebc52"],"state_sha256":"ac3f957502780124413324e8e5830a41884598ba4a9b93a2081afbe6e3a4e47e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"quSfUUrnAd/zr27CcFB6xsSS7SABHZYgT42Spx8PHZ+4CG1lUr9fm5cZbRo0YT5OHDdK55Fbv3RN+chNFVlRAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T11:02:43.240136Z","bundle_sha256":"3f1c6fcb59a5ba8503bb88d81f74079878b59159bc9babb99b11fc18ca15e156"}}