{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:UGLI4NHI2WMOYYGJ2LO35PSYOC","short_pith_number":"pith:UGLI4NHI","schema_version":"1.0","canonical_sha256":"a1968e34e8d598ec60c9d2ddbebe5870a19e93903ee857a3dea41a3071f9aa7c","source":{"kind":"arxiv","id":"2409.10488","version":1},"attestation_state":"computed","paper":{"title":"Do Pre-trained Vision-Language Models Encode Object States?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Chen Sun, David Heffren, Kaleb Newman, Shijie Wang, Yuan Zang","submitted_at":"2024-09-16T17:22:18Z","abstract_excerpt":"For a vision-language model (VLM) to understand the physical world, such as cause and effect, a first step is to capture the temporal dynamics of the visual world, for example how the physical states of objects evolve over time (e.g. a whole apple into a sliced apple). Our paper aims to investigate if VLMs pre-trained on web-scale data learn to encode object states, which can be extracted with zero-shot text prompts. We curate an object state recognition dataset ChangeIt-Frames, and evaluate nine open-source VLMs, including models trained with contrastive and generative objectives. We observe "},"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":"2409.10488","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-09-16T17:22:18Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"97e4b009e5152f1364d1b5969d551fe13b175b54b0c2263c865df83642de333e","abstract_canon_sha256":"e7c8096f20a7f9400239de4a3d45f88d9b811478e717f0bab4d9503287de3cf6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:07:40.954078Z","signature_b64":"ony9bpUHqojDWptiDl/ZqCZ6V1MU3vYoKRcm+79uOfr/pLsuwAjB0zsvo7QD4bolSadmBUzMHqeGDPgbdlPMDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a1968e34e8d598ec60c9d2ddbebe5870a19e93903ee857a3dea41a3071f9aa7c","last_reissued_at":"2026-07-05T09:07:40.953666Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:07:40.953666Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Do Pre-trained Vision-Language Models Encode Object States?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Chen Sun, David Heffren, Kaleb Newman, Shijie Wang, Yuan Zang","submitted_at":"2024-09-16T17:22:18Z","abstract_excerpt":"For a vision-language model (VLM) to understand the physical world, such as cause and effect, a first step is to capture the temporal dynamics of the visual world, for example how the physical states of objects evolve over time (e.g. a whole apple into a sliced apple). Our paper aims to investigate if VLMs pre-trained on web-scale data learn to encode object states, which can be extracted with zero-shot text prompts. We curate an object state recognition dataset ChangeIt-Frames, and evaluate nine open-source VLMs, including models trained with contrastive and generative objectives. We observe "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.10488","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/2409.10488/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":"2409.10488","created_at":"2026-07-05T09:07:40.953721+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.10488v1","created_at":"2026-07-05T09:07:40.953721+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.10488","created_at":"2026-07-05T09:07:40.953721+00:00"},{"alias_kind":"pith_short_12","alias_value":"UGLI4NHI2WMO","created_at":"2026-07-05T09:07:40.953721+00:00"},{"alias_kind":"pith_short_16","alias_value":"UGLI4NHI2WMOYYGJ","created_at":"2026-07-05T09:07:40.953721+00:00"},{"alias_kind":"pith_short_8","alias_value":"UGLI4NHI","created_at":"2026-07-05T09:07:40.953721+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.15130","citing_title":"Enhancing Visual Planning with Auxiliary Tasks and Multi-token Prediction","ref_index":34,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UGLI4NHI2WMOYYGJ2LO35PSYOC","json":"https://pith.science/pith/UGLI4NHI2WMOYYGJ2LO35PSYOC.json","graph_json":"https://pith.science/api/pith-number/UGLI4NHI2WMOYYGJ2LO35PSYOC/graph.json","events_json":"https://pith.science/api/pith-number/UGLI4NHI2WMOYYGJ2LO35PSYOC/events.json","paper":"https://pith.science/paper/UGLI4NHI"},"agent_actions":{"view_html":"https://pith.science/pith/UGLI4NHI2WMOYYGJ2LO35PSYOC","download_json":"https://pith.science/pith/UGLI4NHI2WMOYYGJ2LO35PSYOC.json","view_paper":"https://pith.science/paper/UGLI4NHI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.10488&json=true","fetch_graph":"https://pith.science/api/pith-number/UGLI4NHI2WMOYYGJ2LO35PSYOC/graph.json","fetch_events":"https://pith.science/api/pith-number/UGLI4NHI2WMOYYGJ2LO35PSYOC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UGLI4NHI2WMOYYGJ2LO35PSYOC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UGLI4NHI2WMOYYGJ2LO35PSYOC/action/storage_attestation","attest_author":"https://pith.science/pith/UGLI4NHI2WMOYYGJ2LO35PSYOC/action/author_attestation","sign_citation":"https://pith.science/pith/UGLI4NHI2WMOYYGJ2LO35PSYOC/action/citation_signature","submit_replication":"https://pith.science/pith/UGLI4NHI2WMOYYGJ2LO35PSYOC/action/replication_record"}},"created_at":"2026-07-05T09:07:40.953721+00:00","updated_at":"2026-07-05T09:07:40.953721+00:00"}