{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:HWR5NY4OGEDLKRCZQFF77BHRE4","short_pith_number":"pith:HWR5NY4O","canonical_record":{"source":{"id":"2410.05262","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-07T17:58:47Z","cross_cats_sorted":[],"title_canon_sha256":"24884552b111f1e251a41ae7f02cc238770fecb363f754b79ece546bdeea50b1","abstract_canon_sha256":"9491f211cfa01be07c227d8854d30a5f6eeb1907a0f6c6591516e23d6c6ba315"},"schema_version":"1.0"},"canonical_sha256":"3da3d6e38e3106b54459814bff84f12733477df2c8e7c9b6691724951f5b1ccf","source":{"kind":"arxiv","id":"2410.05262","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.05262","created_at":"2026-07-05T09:17:04Z"},{"alias_kind":"arxiv_version","alias_value":"2410.05262v1","created_at":"2026-07-05T09:17:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.05262","created_at":"2026-07-05T09:17:04Z"},{"alias_kind":"pith_short_12","alias_value":"HWR5NY4OGEDL","created_at":"2026-07-05T09:17:04Z"},{"alias_kind":"pith_short_16","alias_value":"HWR5NY4OGEDLKRCZ","created_at":"2026-07-05T09:17:04Z"},{"alias_kind":"pith_short_8","alias_value":"HWR5NY4O","created_at":"2026-07-05T09:17:04Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:HWR5NY4OGEDLKRCZQFF77BHRE4","target":"record","payload":{"canonical_record":{"source":{"id":"2410.05262","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-07T17:58:47Z","cross_cats_sorted":[],"title_canon_sha256":"24884552b111f1e251a41ae7f02cc238770fecb363f754b79ece546bdeea50b1","abstract_canon_sha256":"9491f211cfa01be07c227d8854d30a5f6eeb1907a0f6c6591516e23d6c6ba315"},"schema_version":"1.0"},"canonical_sha256":"3da3d6e38e3106b54459814bff84f12733477df2c8e7c9b6691724951f5b1ccf","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:17:04.540937Z","signature_b64":"Aw2Q+9lT5IfU79spQ3l5Zz/CEAf+LTOd9VlowZPU2qemI2UH3orwoermbuQzkkJBHFrR05jlpclAauGB8G3lAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3da3d6e38e3106b54459814bff84f12733477df2c8e7c9b6691724951f5b1ccf","last_reissued_at":"2026-07-05T09:17:04.540467Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:17:04.540467Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.05262","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":"2026-07-05T09:17:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mip19nPhVfhnP27VmK0U+gyr5Wx7x20JVRJyBW8TsqGRmB8esbYd+X3fxt9y5eoFgi/sK9B3IkA05EM+dynaCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T02:17:55.166659Z"},"content_sha256":"f364c46e3be349c767b639f3107a41b2f09dd65742cccc373363974862b3a8cd","schema_version":"1.0","event_id":"sha256:f364c46e3be349c767b639f3107a41b2f09dd65742cccc373363974862b3a8cd"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:HWR5NY4OGEDLKRCZQFF77BHRE4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TurtleBench: Evaluating Top Language Models via Real-World Yes/No Puzzles","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hanyu Wang, Ke Fang, Qingchen Yu, Shichao Song, Simin Niu, Yunfeng Shi, Zhiyu Li, Zifan Zheng","submitted_at":"2024-10-07T17:58:47Z","abstract_excerpt":"As the application of Large Language Models (LLMs) expands, the demand for reliable evaluations increases. Existing LLM evaluation benchmarks primarily rely on static datasets, making it challenging to assess model performance in dynamic interactions with users. Moreover, these benchmarks often depend on specific background knowledge, complicating the measurement of a model's logical reasoning capabilities. Other dynamic evaluation methods based on strong models or manual efforts may introduce biases and incur high costs and time demands, hindering large-scale application. To address these iss"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.05262","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/2410.05262/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":"2026-07-05T09:17:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uoSGJYV3FgyYSORQppChYdvAs7UrkmhPnggQfSaW1XONA8hqf4UubhSWnC6L48VCaK+/DiptxuvCf930OlyoDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T02:17:55.167268Z"},"content_sha256":"c8e870411735a042dffd9b1e32360aad7e8b88fb33d49ba35e7c383e4737e06a","schema_version":"1.0","event_id":"sha256:c8e870411735a042dffd9b1e32360aad7e8b88fb33d49ba35e7c383e4737e06a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/HWR5NY4OGEDLKRCZQFF77BHRE4/bundle.json","state_url":"https://pith.science/pith/HWR5NY4OGEDLKRCZQFF77BHRE4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/HWR5NY4OGEDLKRCZQFF77BHRE4/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":"2026-08-05T02:17:55Z","links":{"resolver":"https://pith.science/pith/HWR5NY4OGEDLKRCZQFF77BHRE4","bundle":"https://pith.science/pith/HWR5NY4OGEDLKRCZQFF77BHRE4/bundle.json","state":"https://pith.science/pith/HWR5NY4OGEDLKRCZQFF77BHRE4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/HWR5NY4OGEDLKRCZQFF77BHRE4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:HWR5NY4OGEDLKRCZQFF77BHRE4","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":"9491f211cfa01be07c227d8854d30a5f6eeb1907a0f6c6591516e23d6c6ba315","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-07T17:58:47Z","title_canon_sha256":"24884552b111f1e251a41ae7f02cc238770fecb363f754b79ece546bdeea50b1"},"schema_version":"1.0","source":{"id":"2410.05262","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.05262","created_at":"2026-07-05T09:17:04Z"},{"alias_kind":"arxiv_version","alias_value":"2410.05262v1","created_at":"2026-07-05T09:17:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.05262","created_at":"2026-07-05T09:17:04Z"},{"alias_kind":"pith_short_12","alias_value":"HWR5NY4OGEDL","created_at":"2026-07-05T09:17:04Z"},{"alias_kind":"pith_short_16","alias_value":"HWR5NY4OGEDLKRCZ","created_at":"2026-07-05T09:17:04Z"},{"alias_kind":"pith_short_8","alias_value":"HWR5NY4O","created_at":"2026-07-05T09:17:04Z"}],"graph_snapshots":[{"event_id":"sha256:c8e870411735a042dffd9b1e32360aad7e8b88fb33d49ba35e7c383e4737e06a","target":"graph","created_at":"2026-07-05T09:17:04Z","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/2410.05262/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As the application of Large Language Models (LLMs) expands, the demand for reliable evaluations increases. Existing LLM evaluation benchmarks primarily rely on static datasets, making it challenging to assess model performance in dynamic interactions with users. Moreover, these benchmarks often depend on specific background knowledge, complicating the measurement of a model's logical reasoning capabilities. Other dynamic evaluation methods based on strong models or manual efforts may introduce biases and incur high costs and time demands, hindering large-scale application. To address these iss","authors_text":"Hanyu Wang, Ke Fang, Qingchen Yu, Shichao Song, Simin Niu, Yunfeng Shi, Zhiyu Li, Zifan Zheng","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-07T17:58:47Z","title":"TurtleBench: Evaluating Top Language Models via Real-World Yes/No Puzzles"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.05262","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:f364c46e3be349c767b639f3107a41b2f09dd65742cccc373363974862b3a8cd","target":"record","created_at":"2026-07-05T09:17:04Z","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":"9491f211cfa01be07c227d8854d30a5f6eeb1907a0f6c6591516e23d6c6ba315","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-07T17:58:47Z","title_canon_sha256":"24884552b111f1e251a41ae7f02cc238770fecb363f754b79ece546bdeea50b1"},"schema_version":"1.0","source":{"id":"2410.05262","kind":"arxiv","version":1}},"canonical_sha256":"3da3d6e38e3106b54459814bff84f12733477df2c8e7c9b6691724951f5b1ccf","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3da3d6e38e3106b54459814bff84f12733477df2c8e7c9b6691724951f5b1ccf","first_computed_at":"2026-07-05T09:17:04.540467Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:17:04.540467Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Aw2Q+9lT5IfU79spQ3l5Zz/CEAf+LTOd9VlowZPU2qemI2UH3orwoermbuQzkkJBHFrR05jlpclAauGB8G3lAA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:17:04.540937Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.05262","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f364c46e3be349c767b639f3107a41b2f09dd65742cccc373363974862b3a8cd","sha256:c8e870411735a042dffd9b1e32360aad7e8b88fb33d49ba35e7c383e4737e06a"],"state_sha256":"fc0bcdbd1907c1d25d1f9dc186c5e807b6a3d2d1f3b0bf542a13cadc83afecfd"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lPsYW98jGtKsgSAtq9LYNaI5x/TqHCbDHimeF+qCwCeIdm6O1Gu7RvtDuvpNc1N3fvrOU1hc0cKMXRMVwpJxCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T02:17:55.174186Z","bundle_sha256":"0e7529389176b901644cfda7bbad0d6fb850907bae1da0afb781b3c49a450f06"}}