{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:5NGNKDFUPVM3QPWN4RXAWWLGPE","short_pith_number":"pith:5NGNKDFU","canonical_record":{"source":{"id":"2410.03117","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-10-04T03:21:24Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"e61e91e0d31d51819ae099585367f38ae3dd50f939f8b6a8a74ebb9dcbcb97cd","abstract_canon_sha256":"69d1a679ede03e6038c2e1e3e258b4a9218890e1d347f65b3372b84cf8902d22"},"schema_version":"1.0"},"canonical_sha256":"eb4cd50cb47d59b83ecde46e0b59667928ec7e769ffa324b463e20a8fcfe6375","source":{"kind":"arxiv","id":"2410.03117","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.03117","created_at":"2026-07-05T09:15:44Z"},{"alias_kind":"arxiv_version","alias_value":"2410.03117v1","created_at":"2026-07-05T09:15:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.03117","created_at":"2026-07-05T09:15:44Z"},{"alias_kind":"pith_short_12","alias_value":"5NGNKDFUPVM3","created_at":"2026-07-05T09:15:44Z"},{"alias_kind":"pith_short_16","alias_value":"5NGNKDFUPVM3QPWN","created_at":"2026-07-05T09:15:44Z"},{"alias_kind":"pith_short_8","alias_value":"5NGNKDFU","created_at":"2026-07-05T09:15:44Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:5NGNKDFUPVM3QPWN4RXAWWLGPE","target":"record","payload":{"canonical_record":{"source":{"id":"2410.03117","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-10-04T03:21:24Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"e61e91e0d31d51819ae099585367f38ae3dd50f939f8b6a8a74ebb9dcbcb97cd","abstract_canon_sha256":"69d1a679ede03e6038c2e1e3e258b4a9218890e1d347f65b3372b84cf8902d22"},"schema_version":"1.0"},"canonical_sha256":"eb4cd50cb47d59b83ecde46e0b59667928ec7e769ffa324b463e20a8fcfe6375","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:15:44.726909Z","signature_b64":"4ya0alxU98Gm+Cyr3NuXdRgRX9Sz2LNi7Ob28BLV2BC9BsNSxn4aU8bmaf4eWfqALncIRNj62SFDLcnR488bBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eb4cd50cb47d59b83ecde46e0b59667928ec7e769ffa324b463e20a8fcfe6375","last_reissued_at":"2026-07-05T09:15:44.726409Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:15:44.726409Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.03117","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":"2026-07-05T09:15:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LM0mnfQP6xkMiXWkqfZxuvqzB35gOCXgQHSVjIPser3z5UQZzl/3OJQEXVRjVPzx4KmxkiQrPJQyCCmKvy1dAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T15:25:27.849645Z"},"content_sha256":"550047b771e7bbc46f5e972df2e04a47dbc973bf654cd18d91c55755cf9572b7","schema_version":"1.0","event_id":"sha256:550047b771e7bbc46f5e972df2e04a47dbc973bf654cd18d91c55755cf9572b7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:5NGNKDFUPVM3QPWN4RXAWWLGPE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"ProcBench: Benchmark for Multi-Step Reasoning and Following Procedure","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Hiroki Ikoma, Hiroki Seto, Ippei Fujisawa, Pei-Chun Chien, Rina Onda, Ryota Kanai, Sensho Nobe, Yoshiaki Uchida","submitted_at":"2024-10-04T03:21:24Z","abstract_excerpt":"Reasoning is central to a wide range of intellectual activities, and while the capabilities of large language models (LLMs) continue to advance, their performance in reasoning tasks remains limited. The processes and mechanisms underlying reasoning are not yet fully understood, but key elements include path exploration, selection of relevant knowledge, and multi-step inference. Problems are solved through the synthesis of these components. In this paper, we propose a benchmark that focuses on a specific aspect of reasoning ability: the direct evaluation of multi-step inference. To this end, we"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.03117","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/2410.03117/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":"2026-07-05T09:15:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gWVrSsqluXsP8Ha1pVEEQSx0CZUNIecx6LwqTX+e91gDzYHt1Ic0NKQMSZ5oHsmtag9TpDsx9Fa/vU8gEVbSAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T15:25:27.850179Z"},"content_sha256":"acba0ace870e0c4cf7cd9dcd0aa82ac1b1e84ebd702fdb3778dfbe6f5196143c","schema_version":"1.0","event_id":"sha256:acba0ace870e0c4cf7cd9dcd0aa82ac1b1e84ebd702fdb3778dfbe6f5196143c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/5NGNKDFUPVM3QPWN4RXAWWLGPE/bundle.json","state_url":"https://pith.science/pith/5NGNKDFUPVM3QPWN4RXAWWLGPE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/5NGNKDFUPVM3QPWN4RXAWWLGPE/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":"2026-08-14T15:25:27Z","links":{"resolver":"https://pith.science/pith/5NGNKDFUPVM3QPWN4RXAWWLGPE","bundle":"https://pith.science/pith/5NGNKDFUPVM3QPWN4RXAWWLGPE/bundle.json","state":"https://pith.science/pith/5NGNKDFUPVM3QPWN4RXAWWLGPE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/5NGNKDFUPVM3QPWN4RXAWWLGPE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:5NGNKDFUPVM3QPWN4RXAWWLGPE","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":"69d1a679ede03e6038c2e1e3e258b4a9218890e1d347f65b3372b84cf8902d22","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-10-04T03:21:24Z","title_canon_sha256":"e61e91e0d31d51819ae099585367f38ae3dd50f939f8b6a8a74ebb9dcbcb97cd"},"schema_version":"1.0","source":{"id":"2410.03117","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.03117","created_at":"2026-07-05T09:15:44Z"},{"alias_kind":"arxiv_version","alias_value":"2410.03117v1","created_at":"2026-07-05T09:15:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.03117","created_at":"2026-07-05T09:15:44Z"},{"alias_kind":"pith_short_12","alias_value":"5NGNKDFUPVM3","created_at":"2026-07-05T09:15:44Z"},{"alias_kind":"pith_short_16","alias_value":"5NGNKDFUPVM3QPWN","created_at":"2026-07-05T09:15:44Z"},{"alias_kind":"pith_short_8","alias_value":"5NGNKDFU","created_at":"2026-07-05T09:15:44Z"}],"graph_snapshots":[{"event_id":"sha256:acba0ace870e0c4cf7cd9dcd0aa82ac1b1e84ebd702fdb3778dfbe6f5196143c","target":"graph","created_at":"2026-07-05T09:15:44Z","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/2410.03117/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reasoning is central to a wide range of intellectual activities, and while the capabilities of large language models (LLMs) continue to advance, their performance in reasoning tasks remains limited. The processes and mechanisms underlying reasoning are not yet fully understood, but key elements include path exploration, selection of relevant knowledge, and multi-step inference. Problems are solved through the synthesis of these components. In this paper, we propose a benchmark that focuses on a specific aspect of reasoning ability: the direct evaluation of multi-step inference. To this end, we","authors_text":"Hiroki Ikoma, Hiroki Seto, Ippei Fujisawa, Pei-Chun Chien, Rina Onda, Ryota Kanai, Sensho Nobe, Yoshiaki Uchida","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-10-04T03:21:24Z","title":"ProcBench: Benchmark for Multi-Step Reasoning and Following Procedure"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.03117","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":"sha256:550047b771e7bbc46f5e972df2e04a47dbc973bf654cd18d91c55755cf9572b7","target":"record","created_at":"2026-07-05T09:15:44Z","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":"69d1a679ede03e6038c2e1e3e258b4a9218890e1d347f65b3372b84cf8902d22","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-10-04T03:21:24Z","title_canon_sha256":"e61e91e0d31d51819ae099585367f38ae3dd50f939f8b6a8a74ebb9dcbcb97cd"},"schema_version":"1.0","source":{"id":"2410.03117","kind":"arxiv","version":1}},"canonical_sha256":"eb4cd50cb47d59b83ecde46e0b59667928ec7e769ffa324b463e20a8fcfe6375","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"eb4cd50cb47d59b83ecde46e0b59667928ec7e769ffa324b463e20a8fcfe6375","first_computed_at":"2026-07-05T09:15:44.726409Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:15:44.726409Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4ya0alxU98Gm+Cyr3NuXdRgRX9Sz2LNi7Ob28BLV2BC9BsNSxn4aU8bmaf4eWfqALncIRNj62SFDLcnR488bBg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:15:44.726909Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.03117","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:550047b771e7bbc46f5e972df2e04a47dbc973bf654cd18d91c55755cf9572b7","sha256:acba0ace870e0c4cf7cd9dcd0aa82ac1b1e84ebd702fdb3778dfbe6f5196143c"],"state_sha256":"59e85470c805ad940b2ef7469bc1a3ad1af09c2b3e38b1bec6d24c824c890992"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TfAb8l658NpSvYpOKl4wDphL4rFnexxDLI/JvMKjlSy1uil3Y3nqkxfK4RSYENtLpRAoQOXp0VnRFDIN73N4AA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-14T15:25:27.856557Z","bundle_sha256":"d91e8f77fe3865e6bc5ced9ff5fcab9b670113474f22bd6ea7d3fe73618c04b5"}}