{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:XBFJ4D62JH627W7MPLMZBYTZQW","short_pith_number":"pith:XBFJ4D62","canonical_record":{"source":{"id":"2607.10240","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-07-11T10:01:29Z","cross_cats_sorted":["cs.MM"],"title_canon_sha256":"fc380f1b56fcf5cf2d6a967018edb097f839271905739b7a1b66cfc98ace85d8","abstract_canon_sha256":"60a53b69b3f69bace15e45bb201d587c9a0d3a9bcf1fe6a1e3b876e9cd7c9f65"},"schema_version":"1.0"},"canonical_sha256":"b84a9e0fda49fdafdbec7ad990e27985a068718e72c97f0408256d41dc0ea48c","source":{"kind":"arxiv","id":"2607.10240","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.10240","created_at":"2026-07-14T01:20:32Z"},{"alias_kind":"arxiv_version","alias_value":"2607.10240v1","created_at":"2026-07-14T01:20:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.10240","created_at":"2026-07-14T01:20:32Z"},{"alias_kind":"pith_short_12","alias_value":"XBFJ4D62JH62","created_at":"2026-07-14T01:20:32Z"},{"alias_kind":"pith_short_16","alias_value":"XBFJ4D62JH627W7M","created_at":"2026-07-14T01:20:32Z"},{"alias_kind":"pith_short_8","alias_value":"XBFJ4D62","created_at":"2026-07-14T01:20:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:XBFJ4D62JH627W7MPLMZBYTZQW","target":"record","payload":{"canonical_record":{"source":{"id":"2607.10240","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-07-11T10:01:29Z","cross_cats_sorted":["cs.MM"],"title_canon_sha256":"fc380f1b56fcf5cf2d6a967018edb097f839271905739b7a1b66cfc98ace85d8","abstract_canon_sha256":"60a53b69b3f69bace15e45bb201d587c9a0d3a9bcf1fe6a1e3b876e9cd7c9f65"},"schema_version":"1.0"},"canonical_sha256":"b84a9e0fda49fdafdbec7ad990e27985a068718e72c97f0408256d41dc0ea48c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-14T01:20:32.356001Z","signature_b64":"yKyAndGt9fhfQKgq/REfzCxAABTRIJrjyBVH1fvH7V80Y69bxeHWU604oAWm/jkiGVHGd6zeR6ZHqukSLeH2BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b84a9e0fda49fdafdbec7ad990e27985a068718e72c97f0408256d41dc0ea48c","last_reissued_at":"2026-07-14T01:20:32.355054Z","signature_status":"signed_v1","first_computed_at":"2026-07-14T01:20:32.355054Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.10240","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":"2026-07-14T01:20:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6yKkobbNg7o1Q6vY7H72m+UBEoyYg/1Zw5AdvSLX6ngb0DKzr/IquXq5iiMUgflkOOuwf7rJ76ESzFGpEHNmDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T01:10:54.730200Z"},"content_sha256":"5d3af855f6437adac51dff533f2a9bc2b07db36ae5b4c2cdda1faa11cddaba53","schema_version":"1.0","event_id":"sha256:5d3af855f6437adac51dff533f2a9bc2b07db36ae5b4c2cdda1faa11cddaba53"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:XBFJ4D62JH627W7MPLMZBYTZQW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"What Does Your Short-Answer VQA Score Actually Measure? Evaluator-Dependent Instability in Multimodal Short-Answer Benchmarks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.MM"],"primary_cat":"cs.CV","authors_text":"Guanhua Ye, Meiyu Liang, Niu Jingbin, Yan Li, Yawen Li, Yingxia Shao, Zhe Xue","submitted_at":"2026-07-11T10:01:29Z","abstract_excerpt":"Short-answer VQA benchmarks conflate two distinct quantities: whether a model's answer is semantically correct, and whether that answer matches the surface form expected by the automatic evaluator. We study this conflation across six vision--language models and six benchmarks, using a human-validated semantic judge (97.6% precision) to audit over 37k official errors. A second text-only judge reproduces the same benchmark-level false-negative pattern, showing that the effect is not an artifact of a single audit model. On text-rich benchmarks, up to half of these errors are semantically acceptab"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.10240","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/2607.10240/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":"2026-07-14T01:20:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Wyf6dIVHCOx2XB75nKzh8tzUGQcxWbPD18yTDY1dEn6zpf518HZ7DvCmObxLJbpLf6iaOZgC1m9p6ftkwtMVCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T01:10:54.730702Z"},"content_sha256":"375f3602d4ec94bfdc7529907fd6e6b7fd44554c97b34d0c2ae2b659ccb3e5f1","schema_version":"1.0","event_id":"sha256:375f3602d4ec94bfdc7529907fd6e6b7fd44554c97b34d0c2ae2b659ccb3e5f1"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:XBFJ4D62JH627W7MPLMZBYTZQW","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.18653/v1/2023.emnlp-main.364) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"Ori Yoran, Tomer Wolfson, Ben Bogin, Uri Katz, Daniel S. Weld, and Jonathan Berant. 2023. Answering Questions by Meta- Reasoning over Multiple Chains of Thought. InProceedings of the 2023 Conference on Empirical Methods in Natural Language ","arxiv_id":"2607.10240","detector":"doi_compliance","evidence":{"ref_index":43,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"10.18653/v1/2023.emnlp-","reconstructed_doi":"10.18653/v1/2023.emnlp-main.364"},"severity":"advisory","ref_index":43,"audited_at":"2026-07-29T14:52:04.333112Z","event_type":"pith.integrity.v1","detected_doi":"10.18653/v1/2023.emnlp-main.364","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"09c64722535e23b5360741bb09e9ee95fb08e056609d1044ee16be69e3f6429b","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.1.0","detected_arxiv_id":null,"integrity_event_id":13738,"payload_sha256":"316d2e5518217c16d53a3edbca85da2582325fd653ac39183c399e7e0771389d","signature_b64":null,"signing_key_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-29T14:56:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MDWCambga+T3R89mb1Jq66L86hhMtvRIGopQGB7E5xwNTSqGzMWp5UbdiYeTaaqYoyLtnp4a2bGiaPuVrDldDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T01:10:54.734265Z"},"content_sha256":"8c9997b68176f5de6ad104f8046bd6c5e4e4c87a9e463e02118ed8598650ebbc","schema_version":"1.0","event_id":"sha256:8c9997b68176f5de6ad104f8046bd6c5e4e4c87a9e463e02118ed8598650ebbc"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XBFJ4D62JH627W7MPLMZBYTZQW/bundle.json","state_url":"https://pith.science/pith/XBFJ4D62JH627W7MPLMZBYTZQW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XBFJ4D62JH627W7MPLMZBYTZQW/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":"2026-08-09T01:10:54Z","links":{"resolver":"https://pith.science/pith/XBFJ4D62JH627W7MPLMZBYTZQW","bundle":"https://pith.science/pith/XBFJ4D62JH627W7MPLMZBYTZQW/bundle.json","state":"https://pith.science/pith/XBFJ4D62JH627W7MPLMZBYTZQW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XBFJ4D62JH627W7MPLMZBYTZQW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:XBFJ4D62JH627W7MPLMZBYTZQW","merge_version":"pith-open-graph-merge-v1","event_count":3,"valid_event_count":3,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"60a53b69b3f69bace15e45bb201d587c9a0d3a9bcf1fe6a1e3b876e9cd7c9f65","cross_cats_sorted":["cs.MM"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-07-11T10:01:29Z","title_canon_sha256":"fc380f1b56fcf5cf2d6a967018edb097f839271905739b7a1b66cfc98ace85d8"},"schema_version":"1.0","source":{"id":"2607.10240","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.10240","created_at":"2026-07-14T01:20:32Z"},{"alias_kind":"arxiv_version","alias_value":"2607.10240v1","created_at":"2026-07-14T01:20:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.10240","created_at":"2026-07-14T01:20:32Z"},{"alias_kind":"pith_short_12","alias_value":"XBFJ4D62JH62","created_at":"2026-07-14T01:20:32Z"},{"alias_kind":"pith_short_16","alias_value":"XBFJ4D62JH627W7M","created_at":"2026-07-14T01:20:32Z"},{"alias_kind":"pith_short_8","alias_value":"XBFJ4D62","created_at":"2026-07-14T01:20:32Z"}],"graph_snapshots":[{"event_id":"sha256:375f3602d4ec94bfdc7529907fd6e6b7fd44554c97b34d0c2ae2b659ccb3e5f1","target":"graph","created_at":"2026-07-14T01:20:32Z","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/2607.10240/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Short-answer VQA benchmarks conflate two distinct quantities: whether a model's answer is semantically correct, and whether that answer matches the surface form expected by the automatic evaluator. We study this conflation across six vision--language models and six benchmarks, using a human-validated semantic judge (97.6% precision) to audit over 37k official errors. A second text-only judge reproduces the same benchmark-level false-negative pattern, showing that the effect is not an artifact of a single audit model. On text-rich benchmarks, up to half of these errors are semantically acceptab","authors_text":"Guanhua Ye, Meiyu Liang, Niu Jingbin, Yan Li, Yawen Li, Yingxia Shao, Zhe Xue","cross_cats":["cs.MM"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-07-11T10:01:29Z","title":"What Does Your Short-Answer VQA Score Actually Measure? Evaluator-Dependent Instability in Multimodal Short-Answer Benchmarks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.10240","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":"sha256:5d3af855f6437adac51dff533f2a9bc2b07db36ae5b4c2cdda1faa11cddaba53","target":"record","created_at":"2026-07-14T01:20:32Z","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":"60a53b69b3f69bace15e45bb201d587c9a0d3a9bcf1fe6a1e3b876e9cd7c9f65","cross_cats_sorted":["cs.MM"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-07-11T10:01:29Z","title_canon_sha256":"fc380f1b56fcf5cf2d6a967018edb097f839271905739b7a1b66cfc98ace85d8"},"schema_version":"1.0","source":{"id":"2607.10240","kind":"arxiv","version":1}},"canonical_sha256":"b84a9e0fda49fdafdbec7ad990e27985a068718e72c97f0408256d41dc0ea48c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b84a9e0fda49fdafdbec7ad990e27985a068718e72c97f0408256d41dc0ea48c","first_computed_at":"2026-07-14T01:20:32.355054Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-14T01:20:32.355054Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"yKyAndGt9fhfQKgq/REfzCxAABTRIJrjyBVH1fvH7V80Y69bxeHWU604oAWm/jkiGVHGd6zeR6ZHqukSLeH2BA==","signature_status":"signed_v1","signed_at":"2026-07-14T01:20:32.356001Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.10240","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5d3af855f6437adac51dff533f2a9bc2b07db36ae5b4c2cdda1faa11cddaba53","sha256:375f3602d4ec94bfdc7529907fd6e6b7fd44554c97b34d0c2ae2b659ccb3e5f1","sha256:8c9997b68176f5de6ad104f8046bd6c5e4e4c87a9e463e02118ed8598650ebbc"],"state_sha256":"c401d9128ca87f86f6d5a32faf6ca8faaf930c44984a0e608f3c26bf6a8ba694"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0YN68Ny0zoG3+6Pazwp/EPzumiWfOVjTB/WqNURT+aRn7KwU9v1EGc4n3d8FZ9jvwetTNFBER7GJDhlW/ILVDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T01:10:54.736495Z","bundle_sha256":"340e99354140a6bd0b170148bb9481aaaa624e27ceacf89de42872965a651a42"}}