{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:3VAR6D3MA5DNE7AJ34QK6BDAT5","short_pith_number":"pith:3VAR6D3M","canonical_record":{"source":{"id":"2402.01349","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-02T12:07:00Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"180e07bcb9bc7959d57d62f5676269ad14c26f33e746fef1f9b47819a904be5e","abstract_canon_sha256":"d6ed9dadee32e1024654bb218096e27fc4fcd21b14c283f7c6079401ca85b996"},"schema_version":"1.0"},"canonical_sha256":"dd411f0f6c0746d27c09df20af04609f51b1d9eb74cbacb270e1dcf846b19467","source":{"kind":"arxiv","id":"2402.01349","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.01349","created_at":"2026-07-05T09:45:07Z"},{"alias_kind":"arxiv_version","alias_value":"2402.01349v3","created_at":"2026-07-05T09:45:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.01349","created_at":"2026-07-05T09:45:07Z"},{"alias_kind":"pith_short_12","alias_value":"3VAR6D3MA5DN","created_at":"2026-07-05T09:45:07Z"},{"alias_kind":"pith_short_16","alias_value":"3VAR6D3MA5DNE7AJ","created_at":"2026-07-05T09:45:07Z"},{"alias_kind":"pith_short_8","alias_value":"3VAR6D3M","created_at":"2026-07-05T09:45:07Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:3VAR6D3MA5DNE7AJ34QK6BDAT5","target":"record","payload":{"canonical_record":{"source":{"id":"2402.01349","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-02T12:07:00Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"180e07bcb9bc7959d57d62f5676269ad14c26f33e746fef1f9b47819a904be5e","abstract_canon_sha256":"d6ed9dadee32e1024654bb218096e27fc4fcd21b14c283f7c6079401ca85b996"},"schema_version":"1.0"},"canonical_sha256":"dd411f0f6c0746d27c09df20af04609f51b1d9eb74cbacb270e1dcf846b19467","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:45:07.850644Z","signature_b64":"GQhicp0+wJjnVJm/bbowKj3mifBetfR2zgeHcSHTI3K55jkRhLAlp3FXGNySQIzGfEzwH1ViN+eBQz9SLeLrBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dd411f0f6c0746d27c09df20af04609f51b1d9eb74cbacb270e1dcf846b19467","last_reissued_at":"2026-07-05T09:45:07.850158Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:45:07.850158Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2402.01349","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":"2026-07-05T09:45:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cyvNi/TWVeBmwYQQEMYxZlOOnz4eZXNXgNp3baMjpWX5DVeShKOvC83/GigqRnJAnexChc0YZFuxdfRbHT6EDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T06:27:11.265993Z"},"content_sha256":"ec503bb6e910e4c74c078c6b09dd24ce023873c6607d0f4636e94d70c4b31649","schema_version":"1.0","event_id":"sha256:ec503bb6e910e4c74c078c6b09dd24ce023873c6607d0f4636e94d70c4b31649"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:3VAR6D3MA5DNE7AJ34QK6BDAT5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LLMs May Perform MCQA by Selecting the Least Incorrect Option","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Bing Qin, Haochun Wang, Nuwa Xi, Sendong Zhao, Ting Liu, Zewen Qiang","submitted_at":"2024-02-02T12:07:00Z","abstract_excerpt":"In the field of NLP, Large Language Models (LLMs) have markedly enhanced performance across a variety of tasks. However, the comprehensive evaluation of LLMs remains an inevitable challenge for the community. Recently, the adoption of Multiple Choice Question Answering (MCQA) as a benchmark for assessing LLMs has gained considerable traction. However, concerns regarding the robustness of this evaluative method persist. Building upon previous discussions on the issue of \\textit{variability}, we reveal an additional dimension of concern: LLMs may perform MCQA by selecting the least incorrect opt"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.01349","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/2402.01349/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":"2026-07-05T09:45:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"t/Ob7DTlKKjZSBBHxwNAH/FnExE6oYFBC7QfCKEphIkGQC8HERvamyhfB6XKx6Tvcd2omVakwVnTbhcRyvqICQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T06:27:11.266785Z"},"content_sha256":"4c56ec98699d7b4b02b43c68683d0d27c1a305f40ddc1a0dec1705397a941ae7","schema_version":"1.0","event_id":"sha256:4c56ec98699d7b4b02b43c68683d0d27c1a305f40ddc1a0dec1705397a941ae7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3VAR6D3MA5DNE7AJ34QK6BDAT5/bundle.json","state_url":"https://pith.science/pith/3VAR6D3MA5DNE7AJ34QK6BDAT5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3VAR6D3MA5DNE7AJ34QK6BDAT5/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":"2026-08-06T06:27:11Z","links":{"resolver":"https://pith.science/pith/3VAR6D3MA5DNE7AJ34QK6BDAT5","bundle":"https://pith.science/pith/3VAR6D3MA5DNE7AJ34QK6BDAT5/bundle.json","state":"https://pith.science/pith/3VAR6D3MA5DNE7AJ34QK6BDAT5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3VAR6D3MA5DNE7AJ34QK6BDAT5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:3VAR6D3MA5DNE7AJ34QK6BDAT5","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":"d6ed9dadee32e1024654bb218096e27fc4fcd21b14c283f7c6079401ca85b996","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-02T12:07:00Z","title_canon_sha256":"180e07bcb9bc7959d57d62f5676269ad14c26f33e746fef1f9b47819a904be5e"},"schema_version":"1.0","source":{"id":"2402.01349","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.01349","created_at":"2026-07-05T09:45:07Z"},{"alias_kind":"arxiv_version","alias_value":"2402.01349v3","created_at":"2026-07-05T09:45:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.01349","created_at":"2026-07-05T09:45:07Z"},{"alias_kind":"pith_short_12","alias_value":"3VAR6D3MA5DN","created_at":"2026-07-05T09:45:07Z"},{"alias_kind":"pith_short_16","alias_value":"3VAR6D3MA5DNE7AJ","created_at":"2026-07-05T09:45:07Z"},{"alias_kind":"pith_short_8","alias_value":"3VAR6D3M","created_at":"2026-07-05T09:45:07Z"}],"graph_snapshots":[{"event_id":"sha256:4c56ec98699d7b4b02b43c68683d0d27c1a305f40ddc1a0dec1705397a941ae7","target":"graph","created_at":"2026-07-05T09:45:07Z","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/2402.01349/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In the field of NLP, Large Language Models (LLMs) have markedly enhanced performance across a variety of tasks. However, the comprehensive evaluation of LLMs remains an inevitable challenge for the community. Recently, the adoption of Multiple Choice Question Answering (MCQA) as a benchmark for assessing LLMs has gained considerable traction. However, concerns regarding the robustness of this evaluative method persist. Building upon previous discussions on the issue of \\textit{variability}, we reveal an additional dimension of concern: LLMs may perform MCQA by selecting the least incorrect opt","authors_text":"Bing Qin, Haochun Wang, Nuwa Xi, Sendong Zhao, Ting Liu, Zewen Qiang","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-02T12:07:00Z","title":"LLMs May Perform MCQA by Selecting the Least Incorrect Option"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.01349","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:ec503bb6e910e4c74c078c6b09dd24ce023873c6607d0f4636e94d70c4b31649","target":"record","created_at":"2026-07-05T09:45:07Z","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":"d6ed9dadee32e1024654bb218096e27fc4fcd21b14c283f7c6079401ca85b996","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-02T12:07:00Z","title_canon_sha256":"180e07bcb9bc7959d57d62f5676269ad14c26f33e746fef1f9b47819a904be5e"},"schema_version":"1.0","source":{"id":"2402.01349","kind":"arxiv","version":3}},"canonical_sha256":"dd411f0f6c0746d27c09df20af04609f51b1d9eb74cbacb270e1dcf846b19467","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dd411f0f6c0746d27c09df20af04609f51b1d9eb74cbacb270e1dcf846b19467","first_computed_at":"2026-07-05T09:45:07.850158Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:45:07.850158Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"GQhicp0+wJjnVJm/bbowKj3mifBetfR2zgeHcSHTI3K55jkRhLAlp3FXGNySQIzGfEzwH1ViN+eBQz9SLeLrBg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:45:07.850644Z","signed_message":"canonical_sha256_bytes"},"source_id":"2402.01349","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ec503bb6e910e4c74c078c6b09dd24ce023873c6607d0f4636e94d70c4b31649","sha256:4c56ec98699d7b4b02b43c68683d0d27c1a305f40ddc1a0dec1705397a941ae7"],"state_sha256":"32d553d66d24ed2da7422a5396692f8f8a101f84d4c4075dd2f6cdf82ca4ba29"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"sY3RwJrK1NcHonCt+sNtRoz6LuKv9R2xzh/yt4xu1G6qrzAFgjhNWP+691zfoP7EP3N5xpQL6rHtNGshYzHADA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T06:27:11.274729Z","bundle_sha256":"07f15bc9880fb3bb5478236d943c74eeea2f0c8fb6215586680b89dcf1955ec5"}}