{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:VFRMSIE33TJC7JPQQWIF5PAR2E","short_pith_number":"pith:VFRMSIE3","canonical_record":{"source":{"id":"2305.13160","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-22T15:47:31Z","cross_cats_sorted":[],"title_canon_sha256":"406fe847a07f04446492e26a4760498421a1e1728c63cf003e22dfaa94d6a252","abstract_canon_sha256":"3ad1b4104253d45de5ee8b4238f65e6c8bb0838c9186cde1bef6473df5924aba"},"schema_version":"1.0"},"canonical_sha256":"a962c9209bdcd22fa5f085905ebc11d10d729b12b02f96fdf8893c78ce9df67a","source":{"kind":"arxiv","id":"2305.13160","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.13160","created_at":"2026-07-05T06:59:21Z"},{"alias_kind":"arxiv_version","alias_value":"2305.13160v2","created_at":"2026-07-05T06:59:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.13160","created_at":"2026-07-05T06:59:21Z"},{"alias_kind":"pith_short_12","alias_value":"VFRMSIE33TJC","created_at":"2026-07-05T06:59:21Z"},{"alias_kind":"pith_short_16","alias_value":"VFRMSIE33TJC7JPQ","created_at":"2026-07-05T06:59:21Z"},{"alias_kind":"pith_short_8","alias_value":"VFRMSIE3","created_at":"2026-07-05T06:59:21Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:VFRMSIE33TJC7JPQQWIF5PAR2E","target":"record","payload":{"canonical_record":{"source":{"id":"2305.13160","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-22T15:47:31Z","cross_cats_sorted":[],"title_canon_sha256":"406fe847a07f04446492e26a4760498421a1e1728c63cf003e22dfaa94d6a252","abstract_canon_sha256":"3ad1b4104253d45de5ee8b4238f65e6c8bb0838c9186cde1bef6473df5924aba"},"schema_version":"1.0"},"canonical_sha256":"a962c9209bdcd22fa5f085905ebc11d10d729b12b02f96fdf8893c78ce9df67a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:59:21.277270Z","signature_b64":"1rR1FbUsMKLGQ0m0HhvOmG8iehr/C4BL5KFRrutYEfWfOkDOqJjZflw1iiBByf7ytWVl3uWs+BXADXk84EA0Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a962c9209bdcd22fa5f085905ebc11d10d729b12b02f96fdf8893c78ce9df67a","last_reissued_at":"2026-07-05T06:59:21.276837Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:59:21.276837Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2305.13160","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":"2026-07-05T06:59:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KnbLEJIGUa6TYTNs+iTqZI++sTVROmn9br8GH9CIvMZ723KrPcDfqU3Blp/QZwVmnF5RKLSVZGgytB6xsWX+Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T05:08:18.540415Z"},"content_sha256":"69870b827f47e77b25ca5f7f8ab3e4ca501205dc090e2a26f9f69a014e67c595","schema_version":"1.0","event_id":"sha256:69870b827f47e77b25ca5f7f8ab3e4ca501205dc090e2a26f9f69a014e67c595"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:VFRMSIE33TJC7JPQQWIF5PAR2E","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Can ChatGPT Defend its Belief in Truth? Evaluating LLM Reasoning via Debate","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Boshi Wang, Huan Sun, Xiang Yue","submitted_at":"2023-05-22T15:47:31Z","abstract_excerpt":"Large language models (LLMs) such as ChatGPT and GPT-4 have shown impressive performance in complex reasoning tasks. However, it is difficult to know whether the models are reasoning based on deep understandings of truth and logic, or leveraging their memorized patterns in a relatively superficial way. In this work, we explore testing LLMs' reasoning by engaging with them in a debate-like conversation, where given a question, the LLM and the user need to discuss to make the correct decision starting from opposing arguments. Upon mitigating the Clever Hans effect, our task requires the LLM to n"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.13160","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/2305.13160/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":"2026-07-05T06:59:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1SJMqBw4DV/hwkawhuZImIS76nrBiB2zpZs4aLaQOukUGV7RWEe+KJKxT7toIUcpuhsQGcj70kp1a5BpnrcvBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T05:08:18.541337Z"},"content_sha256":"8358a14953a59d36f6e990af6ca34a269f9079532c6b97d777afadf5f31f334a","schema_version":"1.0","event_id":"sha256:8358a14953a59d36f6e990af6ca34a269f9079532c6b97d777afadf5f31f334a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VFRMSIE33TJC7JPQQWIF5PAR2E/bundle.json","state_url":"https://pith.science/pith/VFRMSIE33TJC7JPQQWIF5PAR2E/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VFRMSIE33TJC7JPQQWIF5PAR2E/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":"2026-08-04T05:08:18Z","links":{"resolver":"https://pith.science/pith/VFRMSIE33TJC7JPQQWIF5PAR2E","bundle":"https://pith.science/pith/VFRMSIE33TJC7JPQQWIF5PAR2E/bundle.json","state":"https://pith.science/pith/VFRMSIE33TJC7JPQQWIF5PAR2E/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VFRMSIE33TJC7JPQQWIF5PAR2E/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:VFRMSIE33TJC7JPQQWIF5PAR2E","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":"3ad1b4104253d45de5ee8b4238f65e6c8bb0838c9186cde1bef6473df5924aba","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-22T15:47:31Z","title_canon_sha256":"406fe847a07f04446492e26a4760498421a1e1728c63cf003e22dfaa94d6a252"},"schema_version":"1.0","source":{"id":"2305.13160","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.13160","created_at":"2026-07-05T06:59:21Z"},{"alias_kind":"arxiv_version","alias_value":"2305.13160v2","created_at":"2026-07-05T06:59:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.13160","created_at":"2026-07-05T06:59:21Z"},{"alias_kind":"pith_short_12","alias_value":"VFRMSIE33TJC","created_at":"2026-07-05T06:59:21Z"},{"alias_kind":"pith_short_16","alias_value":"VFRMSIE33TJC7JPQ","created_at":"2026-07-05T06:59:21Z"},{"alias_kind":"pith_short_8","alias_value":"VFRMSIE3","created_at":"2026-07-05T06:59:21Z"}],"graph_snapshots":[{"event_id":"sha256:8358a14953a59d36f6e990af6ca34a269f9079532c6b97d777afadf5f31f334a","target":"graph","created_at":"2026-07-05T06:59:21Z","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/2305.13160/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) such as ChatGPT and GPT-4 have shown impressive performance in complex reasoning tasks. However, it is difficult to know whether the models are reasoning based on deep understandings of truth and logic, or leveraging their memorized patterns in a relatively superficial way. In this work, we explore testing LLMs' reasoning by engaging with them in a debate-like conversation, where given a question, the LLM and the user need to discuss to make the correct decision starting from opposing arguments. Upon mitigating the Clever Hans effect, our task requires the LLM to n","authors_text":"Boshi Wang, Huan Sun, Xiang Yue","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-22T15:47:31Z","title":"Can ChatGPT Defend its Belief in Truth? Evaluating LLM Reasoning via Debate"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.13160","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":"sha256:69870b827f47e77b25ca5f7f8ab3e4ca501205dc090e2a26f9f69a014e67c595","target":"record","created_at":"2026-07-05T06:59:21Z","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":"3ad1b4104253d45de5ee8b4238f65e6c8bb0838c9186cde1bef6473df5924aba","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-22T15:47:31Z","title_canon_sha256":"406fe847a07f04446492e26a4760498421a1e1728c63cf003e22dfaa94d6a252"},"schema_version":"1.0","source":{"id":"2305.13160","kind":"arxiv","version":2}},"canonical_sha256":"a962c9209bdcd22fa5f085905ebc11d10d729b12b02f96fdf8893c78ce9df67a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a962c9209bdcd22fa5f085905ebc11d10d729b12b02f96fdf8893c78ce9df67a","first_computed_at":"2026-07-05T06:59:21.276837Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:59:21.276837Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"1rR1FbUsMKLGQ0m0HhvOmG8iehr/C4BL5KFRrutYEfWfOkDOqJjZflw1iiBByf7ytWVl3uWs+BXADXk84EA0Ag==","signature_status":"signed_v1","signed_at":"2026-07-05T06:59:21.277270Z","signed_message":"canonical_sha256_bytes"},"source_id":"2305.13160","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:69870b827f47e77b25ca5f7f8ab3e4ca501205dc090e2a26f9f69a014e67c595","sha256:8358a14953a59d36f6e990af6ca34a269f9079532c6b97d777afadf5f31f334a"],"state_sha256":"15104e3134eb84acb11144bf8c041e134a44c0284d9ef854e3a7e330350c5adb"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NhavE2eG495yZcG5MFdZ9GvCgsvo6kh3+9Vt3kF5rovJ8nmplemoRcAsh+JB9IFUGnQUO/jugtVUHnuGrbMiAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T05:08:18.548684Z","bundle_sha256":"114c8383c0fdc02c27272f2fa568424b369014eca26b42e6bf738b13bb02ce9c"}}