{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:W5VGGY53FL2PH7ZBLOGURPCKL2","short_pith_number":"pith:W5VGGY53","schema_version":"1.0","canonical_sha256":"b76a6363bb2af4f3ff215b8d48bc4a5eaf4a8e5f90715cb6c25998e9f116e5fe","source":{"kind":"arxiv","id":"2502.02740","version":1},"attestation_state":"computed","paper":{"title":"Vision-Language Model Dialog Games for Self-Improvement","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Christos Kaplanis, Ksenia Konyushkova, Misha Denil, Serkan Cabi","submitted_at":"2025-02-04T21:58:07Z","abstract_excerpt":"The increasing demand for high-quality, diverse training data poses a significant bottleneck in advancing vision-language models (VLMs). This paper presents VLM Dialog Games, a novel and scalable self-improvement framework for VLMs. Our approach leverages self-play between two agents engaged in a goal-oriented play centered around image identification. By filtering for successful game interactions, we automatically curate a high-quality dataset of interleaved images and text. We demonstrate that fine-tuning on this synthetic data leads to performance gains on downstream tasks and generalises a"},"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":"2502.02740","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-04T21:58:07Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7e1c44c2b953e6480e3b2b5ad3ba03f66b68f37bc5f9486917ca71bce1c3cc0c","abstract_canon_sha256":"72f2dea3d6497ad4a81382dc1244434dbf29c55d7794c0392ab89448e976173f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:09:54.513342Z","signature_b64":"nufWLKL39r/LREQ0qU3fRB4vfpvYuuV7WVn9KKf1D5tqL4tF6l4JTHe2+I638EIAnI+QZYyKoUobDqiOXEjrAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b76a6363bb2af4f3ff215b8d48bc4a5eaf4a8e5f90715cb6c25998e9f116e5fe","last_reissued_at":"2026-07-05T10:09:54.512986Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:09:54.512986Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Vision-Language Model Dialog Games for Self-Improvement","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Christos Kaplanis, Ksenia Konyushkova, Misha Denil, Serkan Cabi","submitted_at":"2025-02-04T21:58:07Z","abstract_excerpt":"The increasing demand for high-quality, diverse training data poses a significant bottleneck in advancing vision-language models (VLMs). This paper presents VLM Dialog Games, a novel and scalable self-improvement framework for VLMs. Our approach leverages self-play between two agents engaged in a goal-oriented play centered around image identification. By filtering for successful game interactions, we automatically curate a high-quality dataset of interleaved images and text. We demonstrate that fine-tuning on this synthetic data leads to performance gains on downstream tasks and generalises a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.02740","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/2502.02740/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":"2502.02740","created_at":"2026-07-05T10:09:54.513041+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.02740v1","created_at":"2026-07-05T10:09:54.513041+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.02740","created_at":"2026-07-05T10:09:54.513041+00:00"},{"alias_kind":"pith_short_12","alias_value":"W5VGGY53FL2P","created_at":"2026-07-05T10:09:54.513041+00:00"},{"alias_kind":"pith_short_16","alias_value":"W5VGGY53FL2PH7ZB","created_at":"2026-07-05T10:09:54.513041+00:00"},{"alias_kind":"pith_short_8","alias_value":"W5VGGY53","created_at":"2026-07-05T10:09:54.513041+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/W5VGGY53FL2PH7ZBLOGURPCKL2","json":"https://pith.science/pith/W5VGGY53FL2PH7ZBLOGURPCKL2.json","graph_json":"https://pith.science/api/pith-number/W5VGGY53FL2PH7ZBLOGURPCKL2/graph.json","events_json":"https://pith.science/api/pith-number/W5VGGY53FL2PH7ZBLOGURPCKL2/events.json","paper":"https://pith.science/paper/W5VGGY53"},"agent_actions":{"view_html":"https://pith.science/pith/W5VGGY53FL2PH7ZBLOGURPCKL2","download_json":"https://pith.science/pith/W5VGGY53FL2PH7ZBLOGURPCKL2.json","view_paper":"https://pith.science/paper/W5VGGY53","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.02740&json=true","fetch_graph":"https://pith.science/api/pith-number/W5VGGY53FL2PH7ZBLOGURPCKL2/graph.json","fetch_events":"https://pith.science/api/pith-number/W5VGGY53FL2PH7ZBLOGURPCKL2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/W5VGGY53FL2PH7ZBLOGURPCKL2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/W5VGGY53FL2PH7ZBLOGURPCKL2/action/storage_attestation","attest_author":"https://pith.science/pith/W5VGGY53FL2PH7ZBLOGURPCKL2/action/author_attestation","sign_citation":"https://pith.science/pith/W5VGGY53FL2PH7ZBLOGURPCKL2/action/citation_signature","submit_replication":"https://pith.science/pith/W5VGGY53FL2PH7ZBLOGURPCKL2/action/replication_record"}},"created_at":"2026-07-05T10:09:54.513041+00:00","updated_at":"2026-07-05T10:09:54.513041+00:00"}