{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:MCSKYJ3UOWTQKFSOOUOUOYETAB","short_pith_number":"pith:MCSKYJ3U","schema_version":"1.0","canonical_sha256":"60a4ac277475a705164e751d47609300744685f40f60325abae6e2292bdfc18e","source":{"kind":"arxiv","id":"2607.16311","version":1},"attestation_state":"computed","paper":{"title":"Seeing What Is Actually There: PriVE-Bench and PriVE-Tools for Counterfactual Evaluation of Agentic Visual Evidence in VLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Guoyi Xu, Hongpeng Zhou, Jiachen Tu, Jingyuan Sun, Jingyu Sun, Rui Qian, Yan Lin, Yaoxin Jiang, Yizheng Sun, Yuyang Xue, Zhengtao Yao","submitted_at":"2026-07-14T16:45:02Z","abstract_excerpt":"Vision-language models (VLMs) often answer visual questions using learned language and category priors rather than grounding their predictions in the image itself. Counterfactual images provide a natural diagnostic setting for this failure mode: when visible evidence contradicts what is usually true, a grounded model should answer from the pixels, while a prior-following model will produce a canonical but visually incorrect response. However, existing counterfactual benchmarks mainly ask whether such prior-following behavior exists. In this paper, we ask a further question motivated by the ris"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2607.16311","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-07-14T16:45:02Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ea7cf71cc86aaf0d88ccc0f859bb709fbdd04a7d3798f130a3658ba861840216","abstract_canon_sha256":"6dde087ec3160f6f835e7535b94a47a3943a4525f3206b88077aeebc8105f9ac"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-21T00:20:11.690256Z","signature_b64":"P9MS5KzeUnwplyxbFa9JSycaOih/MkJaaTIX1rHHRHHlUpCW7eBRI+bR8KmQ5vfkUwibEYYUbpEtmxLCMc8QDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"60a4ac277475a705164e751d47609300744685f40f60325abae6e2292bdfc18e","last_reissued_at":"2026-07-21T00:20:11.689274Z","signature_status":"signed_v1","first_computed_at":"2026-07-21T00:20:11.689274Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Seeing What Is Actually There: PriVE-Bench and PriVE-Tools for Counterfactual Evaluation of Agentic Visual Evidence in VLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Guoyi Xu, Hongpeng Zhou, Jiachen Tu, Jingyuan Sun, Jingyu Sun, Rui Qian, Yan Lin, Yaoxin Jiang, Yizheng Sun, Yuyang Xue, Zhengtao Yao","submitted_at":"2026-07-14T16:45:02Z","abstract_excerpt":"Vision-language models (VLMs) often answer visual questions using learned language and category priors rather than grounding their predictions in the image itself. Counterfactual images provide a natural diagnostic setting for this failure mode: when visible evidence contradicts what is usually true, a grounded model should answer from the pixels, while a prior-following model will produce a canonical but visually incorrect response. However, existing counterfactual benchmarks mainly ask whether such prior-following behavior exists. In this paper, we ask a further question motivated by the ris"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.16311","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/2607.16311/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2607.16311","created_at":"2026-07-21T00:20:11.689764+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.16311v1","created_at":"2026-07-21T00:20:11.689764+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.16311","created_at":"2026-07-21T00:20:11.689764+00:00"},{"alias_kind":"pith_short_12","alias_value":"MCSKYJ3UOWTQ","created_at":"2026-07-21T00:20:11.689764+00:00"},{"alias_kind":"pith_short_16","alias_value":"MCSKYJ3UOWTQKFSO","created_at":"2026-07-21T00:20:11.689764+00:00"},{"alias_kind":"pith_short_8","alias_value":"MCSKYJ3U","created_at":"2026-07-21T00:20:11.689764+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MCSKYJ3UOWTQKFSOOUOUOYETAB","json":"https://pith.science/pith/MCSKYJ3UOWTQKFSOOUOUOYETAB.json","graph_json":"https://pith.science/api/pith-number/MCSKYJ3UOWTQKFSOOUOUOYETAB/graph.json","events_json":"https://pith.science/api/pith-number/MCSKYJ3UOWTQKFSOOUOUOYETAB/events.json","paper":"https://pith.science/paper/MCSKYJ3U"},"agent_actions":{"view_html":"https://pith.science/pith/MCSKYJ3UOWTQKFSOOUOUOYETAB","download_json":"https://pith.science/pith/MCSKYJ3UOWTQKFSOOUOUOYETAB.json","view_paper":"https://pith.science/paper/MCSKYJ3U","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.16311&json=true","fetch_graph":"https://pith.science/api/pith-number/MCSKYJ3UOWTQKFSOOUOUOYETAB/graph.json","fetch_events":"https://pith.science/api/pith-number/MCSKYJ3UOWTQKFSOOUOUOYETAB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MCSKYJ3UOWTQKFSOOUOUOYETAB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MCSKYJ3UOWTQKFSOOUOUOYETAB/action/storage_attestation","attest_author":"https://pith.science/pith/MCSKYJ3UOWTQKFSOOUOUOYETAB/action/author_attestation","sign_citation":"https://pith.science/pith/MCSKYJ3UOWTQKFSOOUOUOYETAB/action/citation_signature","submit_replication":"https://pith.science/pith/MCSKYJ3UOWTQKFSOOUOUOYETAB/action/replication_record"}},"created_at":"2026-07-21T00:20:11.689764+00:00","updated_at":"2026-07-21T00:20:11.689764+00:00"}