{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:LCVXSKXLRT26GNBTTIEJCS6CCA","short_pith_number":"pith:LCVXSKXL","schema_version":"1.0","canonical_sha256":"58ab792aeb8cf5e334339a08914bc210232e43bbe5d657da5c03174db65f594e","source":{"kind":"arxiv","id":"2501.04568","version":2},"attestation_state":"computed","paper":{"title":"Feedback-Driven Vision-Language Alignment with Minimal Human Supervision","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Aleix Martinez, Evgeny Perevodchikov, Giorgio Giannone, Qianli Feng, Rui Chen, Ruoteng Li","submitted_at":"2025-01-08T15:32:12Z","abstract_excerpt":"Vision-language models (VLMs) have demonstrated remarkable potential in integrating visual and linguistic information, but their performance is often constrained by the need for extensive, high-quality image-text training data. Curation of these image-text pairs is both time-consuming and computationally expensive. To address this challenge, we introduce SVP (Sampling-based Visual Projection), a novel framework that enhances vision-language alignment without relying on manually curated text-image pairs or preference annotation. SVP leverages a small set of manually selected images, self-captio"},"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":"2501.04568","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-01-08T15:32:12Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"title_canon_sha256":"e836f6032d6cc774a2b629e1f63e64f7616e4e3ee6af8ce979920ceb190891aa","abstract_canon_sha256":"0480f58ffe5691a2d4d62a06f6bdf150a8e2cc38e6884cd6bcf1ae45717f1ed6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:05:22.879263Z","signature_b64":"xSzXOegWt5G0FZpkYXupB7Y+Aac0TQgAOmwbGJuDVPNGwrjD6UwBC4zcFngydqiTMIo2lqzv+t39JiHz5lRJBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"58ab792aeb8cf5e334339a08914bc210232e43bbe5d657da5c03174db65f594e","last_reissued_at":"2026-07-05T11:05:22.878777Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:05:22.878777Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Feedback-Driven Vision-Language Alignment with Minimal Human Supervision","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Aleix Martinez, Evgeny Perevodchikov, Giorgio Giannone, Qianli Feng, Rui Chen, Ruoteng Li","submitted_at":"2025-01-08T15:32:12Z","abstract_excerpt":"Vision-language models (VLMs) have demonstrated remarkable potential in integrating visual and linguistic information, but their performance is often constrained by the need for extensive, high-quality image-text training data. Curation of these image-text pairs is both time-consuming and computationally expensive. To address this challenge, we introduce SVP (Sampling-based Visual Projection), a novel framework that enhances vision-language alignment without relying on manually curated text-image pairs or preference annotation. SVP leverages a small set of manually selected images, self-captio"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.04568","kind":"arxiv","version":2},"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/2501.04568/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":"2501.04568","created_at":"2026-07-05T11:05:22.878839+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.04568v2","created_at":"2026-07-05T11:05:22.878839+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.04568","created_at":"2026-07-05T11:05:22.878839+00:00"},{"alias_kind":"pith_short_12","alias_value":"LCVXSKXLRT26","created_at":"2026-07-05T11:05:22.878839+00:00"},{"alias_kind":"pith_short_16","alias_value":"LCVXSKXLRT26GNBT","created_at":"2026-07-05T11:05:22.878839+00:00"},{"alias_kind":"pith_short_8","alias_value":"LCVXSKXL","created_at":"2026-07-05T11:05:22.878839+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.05614","citing_title":"Grounding Hierarchical Vision-Language-Action Models Through Explicit Language-Action Alignment","ref_index":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LCVXSKXLRT26GNBTTIEJCS6CCA","json":"https://pith.science/pith/LCVXSKXLRT26GNBTTIEJCS6CCA.json","graph_json":"https://pith.science/api/pith-number/LCVXSKXLRT26GNBTTIEJCS6CCA/graph.json","events_json":"https://pith.science/api/pith-number/LCVXSKXLRT26GNBTTIEJCS6CCA/events.json","paper":"https://pith.science/paper/LCVXSKXL"},"agent_actions":{"view_html":"https://pith.science/pith/LCVXSKXLRT26GNBTTIEJCS6CCA","download_json":"https://pith.science/pith/LCVXSKXLRT26GNBTTIEJCS6CCA.json","view_paper":"https://pith.science/paper/LCVXSKXL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.04568&json=true","fetch_graph":"https://pith.science/api/pith-number/LCVXSKXLRT26GNBTTIEJCS6CCA/graph.json","fetch_events":"https://pith.science/api/pith-number/LCVXSKXLRT26GNBTTIEJCS6CCA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LCVXSKXLRT26GNBTTIEJCS6CCA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LCVXSKXLRT26GNBTTIEJCS6CCA/action/storage_attestation","attest_author":"https://pith.science/pith/LCVXSKXLRT26GNBTTIEJCS6CCA/action/author_attestation","sign_citation":"https://pith.science/pith/LCVXSKXLRT26GNBTTIEJCS6CCA/action/citation_signature","submit_replication":"https://pith.science/pith/LCVXSKXLRT26GNBTTIEJCS6CCA/action/replication_record"}},"created_at":"2026-07-05T11:05:22.878839+00:00","updated_at":"2026-07-05T11:05:22.878839+00:00"}