{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:VNYVMYQOWBNZUE63BRRXAF3RHS","short_pith_number":"pith:VNYVMYQO","canonical_record":{"source":{"id":"2305.15269","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T15:55:51Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"603c297e9e90d1ecc76882fae5041b7492ddc5b12bd1d642d18b3b86347aab84","abstract_canon_sha256":"2a3eb86f9b23fcb8f18d9984afaf8dbf03f7398fc1c10fae4403290ba879e9a4"},"schema_version":"1.0"},"canonical_sha256":"ab7156620eb05b9a13db0c637017713cb58169426474de5d918aa030e060c284","source":{"kind":"arxiv","id":"2305.15269","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.15269","created_at":"2026-07-05T07:08:58Z"},{"alias_kind":"arxiv_version","alias_value":"2305.15269v3","created_at":"2026-07-05T07:08:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.15269","created_at":"2026-07-05T07:08:58Z"},{"alias_kind":"pith_short_12","alias_value":"VNYVMYQOWBNZ","created_at":"2026-07-05T07:08:58Z"},{"alias_kind":"pith_short_16","alias_value":"VNYVMYQOWBNZUE63","created_at":"2026-07-05T07:08:58Z"},{"alias_kind":"pith_short_8","alias_value":"VNYVMYQO","created_at":"2026-07-05T07:08:58Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:VNYVMYQOWBNZUE63BRRXAF3RHS","target":"record","payload":{"canonical_record":{"source":{"id":"2305.15269","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T15:55:51Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"603c297e9e90d1ecc76882fae5041b7492ddc5b12bd1d642d18b3b86347aab84","abstract_canon_sha256":"2a3eb86f9b23fcb8f18d9984afaf8dbf03f7398fc1c10fae4403290ba879e9a4"},"schema_version":"1.0"},"canonical_sha256":"ab7156620eb05b9a13db0c637017713cb58169426474de5d918aa030e060c284","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:08:58.103489Z","signature_b64":"ihBcHjUvUGYr4wCwhPgD8KfmA2RtrdPMnBsTVD0vfyelmxHYWKsb6vK3UIPAC6KG94JNLo8WlPgrrzgnGN24Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ab7156620eb05b9a13db0c637017713cb58169426474de5d918aa030e060c284","last_reissued_at":"2026-07-05T07:08:58.102986Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:08:58.102986Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2305.15269","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:08:58Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SxZHlF17nGwc2QClnYslWBS7L5zIs7Bow5nCLZ/LF8qkzW6mrGqcADXdtV2tf7WWYP0pmnobH21mSlabLmyDCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-19T02:25:54.603466Z"},"content_sha256":"74dbfbfb08da59b968806c2afa5e967ad4762a8f2dd8954c811016636fb9840e","schema_version":"1.0","event_id":"sha256:74dbfbfb08da59b968806c2afa5e967ad4762a8f2dd8954c811016636fb9840e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:VNYVMYQOWBNZUE63BRRXAF3RHS","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Testing the General Deductive Reasoning Capacity of Large Language Models Using OOD Examples","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Abulhair Saparov, He He, Najoung Kim, Nitish Joshi, Richard Yuanzhe Pang, Seyed Mehran Kazemi, Vishakh Padmakumar","submitted_at":"2023-05-24T15:55:51Z","abstract_excerpt":"Given the intractably large size of the space of proofs, any model that is capable of general deductive reasoning must generalize to proofs of greater complexity. Recent studies have shown that large language models (LLMs) possess some abstract deductive reasoning ability given chain-of-thought prompts. However, they have primarily been tested on proofs using modus ponens or of a specific size, and from the same distribution as the in-context examples. To measure the general deductive reasoning ability of LLMs, we test on a broad set of deduction rules and measure their ability to generalize t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.15269","kind":"arxiv","version":3},"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/2305.15269/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":"2026-07-05T07:08:58Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"x6cooU5mv5Ibt7GV34aCEydp8F8wOTT8987flS78QFsi37KhnL3iORev/4Ws6jWTVFU69+zcm+T5aNmNXBmMDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-19T02:25:54.603843Z"},"content_sha256":"fb7b347b7f7338d3af4613ce771c90da3180a8a549ce4a7f6a99ef56cb7d8198","schema_version":"1.0","event_id":"sha256:fb7b347b7f7338d3af4613ce771c90da3180a8a549ce4a7f6a99ef56cb7d8198"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VNYVMYQOWBNZUE63BRRXAF3RHS/bundle.json","state_url":"https://pith.science/pith/VNYVMYQOWBNZUE63BRRXAF3RHS/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VNYVMYQOWBNZUE63BRRXAF3RHS/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":"2026-07-19T02:25:54Z","links":{"resolver":"https://pith.science/pith/VNYVMYQOWBNZUE63BRRXAF3RHS","bundle":"https://pith.science/pith/VNYVMYQOWBNZUE63BRRXAF3RHS/bundle.json","state":"https://pith.science/pith/VNYVMYQOWBNZUE63BRRXAF3RHS/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VNYVMYQOWBNZUE63BRRXAF3RHS/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:VNYVMYQOWBNZUE63BRRXAF3RHS","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":"2a3eb86f9b23fcb8f18d9984afaf8dbf03f7398fc1c10fae4403290ba879e9a4","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T15:55:51Z","title_canon_sha256":"603c297e9e90d1ecc76882fae5041b7492ddc5b12bd1d642d18b3b86347aab84"},"schema_version":"1.0","source":{"id":"2305.15269","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.15269","created_at":"2026-07-05T07:08:58Z"},{"alias_kind":"arxiv_version","alias_value":"2305.15269v3","created_at":"2026-07-05T07:08:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.15269","created_at":"2026-07-05T07:08:58Z"},{"alias_kind":"pith_short_12","alias_value":"VNYVMYQOWBNZ","created_at":"2026-07-05T07:08:58Z"},{"alias_kind":"pith_short_16","alias_value":"VNYVMYQOWBNZUE63","created_at":"2026-07-05T07:08:58Z"},{"alias_kind":"pith_short_8","alias_value":"VNYVMYQO","created_at":"2026-07-05T07:08:58Z"}],"graph_snapshots":[{"event_id":"sha256:fb7b347b7f7338d3af4613ce771c90da3180a8a549ce4a7f6a99ef56cb7d8198","target":"graph","created_at":"2026-07-05T07:08:58Z","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/2305.15269/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Given the intractably large size of the space of proofs, any model that is capable of general deductive reasoning must generalize to proofs of greater complexity. Recent studies have shown that large language models (LLMs) possess some abstract deductive reasoning ability given chain-of-thought prompts. However, they have primarily been tested on proofs using modus ponens or of a specific size, and from the same distribution as the in-context examples. To measure the general deductive reasoning ability of LLMs, we test on a broad set of deduction rules and measure their ability to generalize t","authors_text":"Abulhair Saparov, He He, Najoung Kim, Nitish Joshi, Richard Yuanzhe Pang, Seyed Mehran Kazemi, Vishakh Padmakumar","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T15:55:51Z","title":"Testing the General Deductive Reasoning Capacity of Large Language Models Using OOD Examples"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.15269","kind":"arxiv","version":3},"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":"sha256:74dbfbfb08da59b968806c2afa5e967ad4762a8f2dd8954c811016636fb9840e","target":"record","created_at":"2026-07-05T07:08:58Z","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":"2a3eb86f9b23fcb8f18d9984afaf8dbf03f7398fc1c10fae4403290ba879e9a4","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T15:55:51Z","title_canon_sha256":"603c297e9e90d1ecc76882fae5041b7492ddc5b12bd1d642d18b3b86347aab84"},"schema_version":"1.0","source":{"id":"2305.15269","kind":"arxiv","version":3}},"canonical_sha256":"ab7156620eb05b9a13db0c637017713cb58169426474de5d918aa030e060c284","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ab7156620eb05b9a13db0c637017713cb58169426474de5d918aa030e060c284","first_computed_at":"2026-07-05T07:08:58.102986Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:08:58.102986Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ihBcHjUvUGYr4wCwhPgD8KfmA2RtrdPMnBsTVD0vfyelmxHYWKsb6vK3UIPAC6KG94JNLo8WlPgrrzgnGN24Cw==","signature_status":"signed_v1","signed_at":"2026-07-05T07:08:58.103489Z","signed_message":"canonical_sha256_bytes"},"source_id":"2305.15269","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:74dbfbfb08da59b968806c2afa5e967ad4762a8f2dd8954c811016636fb9840e","sha256:fb7b347b7f7338d3af4613ce771c90da3180a8a549ce4a7f6a99ef56cb7d8198"],"state_sha256":"9b7dfb51b1b871ed761e1a4fed4370fe4e064a25521f2893839ba3fb6ef209ee"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dGptPSEEVEIJgEMA414AoenRq4B745HtRbdtsC8msQJm9UiylkGS5pbSTpl87/EQoyvYbpaYlaNTz+krmAeTBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-19T02:25:54.606409Z","bundle_sha256":"42c16505b1faa1aff7bfbd0ead892a1b264600f1a67d62401a1212ee99084632"}}