{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:A6G7TRL7SKK57KCSH27POC4ETH","short_pith_number":"pith:A6G7TRL7","schema_version":"1.0","canonical_sha256":"078df9c57f9295dfa8523ebef70b8499d54d758741462ef6441c6cf243516a2b","source":{"kind":"arxiv","id":"2511.11440","version":3},"attestation_state":"computed","paper":{"title":"Synthetic Stimuli, Real Gains: Rethinking VLM Fine-Tuning Through Fully Controlled Data Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Giuseppe Riccardi, Massimo Rizzoli, Seyed Mahed Mousavi, Simone Alghisi","submitted_at":"2025-11-14T16:07:18Z","abstract_excerpt":"Performance gains of Vision Language Models (VLMs) obtained by fine-tuning are generally based on ad hoc data collection and annotation of real-world scenes. Despite the improvements, this process is often prone to biases, errors, and distribution imbalance, resulting in overfitting and imbalanced performance. Although a few studies have explored synthetic data generation, they typically lack control over data distribution and annotation quality. In this work, we re-evaluate the potential of model fine-tuning by exploring a fully controlled data generation and annotation pipeline, obtaining bi"},"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":"2511.11440","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-11-14T16:07:18Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"eca200f0bf3800888181f69ed229af3f6c63e56095403892d3944c9d4a687269","abstract_canon_sha256":"72634096b3a174a96c2b35a3656da4b761c5f52a5ca98c8ceee814685ab13c90"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T01:03:46.562783Z","signature_b64":"lidZyLZNEj0QxgxgoH1hWYVjK4hTHME1rqugojJofETM4hcSMmaPoYAtCBTkt3/8wEGZOLyuOQf4XMnFek9GAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"078df9c57f9295dfa8523ebef70b8499d54d758741462ef6441c6cf243516a2b","last_reissued_at":"2026-06-01T01:03:46.561767Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T01:03:46.561767Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Synthetic Stimuli, Real Gains: Rethinking VLM Fine-Tuning Through Fully Controlled Data Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Giuseppe Riccardi, Massimo Rizzoli, Seyed Mahed Mousavi, Simone Alghisi","submitted_at":"2025-11-14T16:07:18Z","abstract_excerpt":"Performance gains of Vision Language Models (VLMs) obtained by fine-tuning are generally based on ad hoc data collection and annotation of real-world scenes. Despite the improvements, this process is often prone to biases, errors, and distribution imbalance, resulting in overfitting and imbalanced performance. Although a few studies have explored synthetic data generation, they typically lack control over data distribution and annotation quality. In this work, we re-evaluate the potential of model fine-tuning by exploring a fully controlled data generation and annotation pipeline, obtaining bi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.11440","kind":"arxiv","version":3},"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/2511.11440/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":"2511.11440","created_at":"2026-06-01T01:03:46.561925+00:00"},{"alias_kind":"arxiv_version","alias_value":"2511.11440v3","created_at":"2026-06-01T01:03:46.561925+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.11440","created_at":"2026-06-01T01:03:46.561925+00:00"},{"alias_kind":"pith_short_12","alias_value":"A6G7TRL7SKK5","created_at":"2026-06-01T01:03:46.561925+00:00"},{"alias_kind":"pith_short_16","alias_value":"A6G7TRL7SKK57KCS","created_at":"2026-06-01T01:03:46.561925+00:00"},{"alias_kind":"pith_short_8","alias_value":"A6G7TRL7","created_at":"2026-06-01T01:03:46.561925+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2604.02812","citing_title":"Learning Structured Robot Policies from Vision-Language Models via Synthetic Neuro-Symbolic Supervision","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"2604.02812","citing_title":"Learning Structured Robot Policies from Vision-Language Models via Synthetic Neuro-Symbolic Supervision","ref_index":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/A6G7TRL7SKK57KCSH27POC4ETH","json":"https://pith.science/pith/A6G7TRL7SKK57KCSH27POC4ETH.json","graph_json":"https://pith.science/api/pith-number/A6G7TRL7SKK57KCSH27POC4ETH/graph.json","events_json":"https://pith.science/api/pith-number/A6G7TRL7SKK57KCSH27POC4ETH/events.json","paper":"https://pith.science/paper/A6G7TRL7"},"agent_actions":{"view_html":"https://pith.science/pith/A6G7TRL7SKK57KCSH27POC4ETH","download_json":"https://pith.science/pith/A6G7TRL7SKK57KCSH27POC4ETH.json","view_paper":"https://pith.science/paper/A6G7TRL7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2511.11440&json=true","fetch_graph":"https://pith.science/api/pith-number/A6G7TRL7SKK57KCSH27POC4ETH/graph.json","fetch_events":"https://pith.science/api/pith-number/A6G7TRL7SKK57KCSH27POC4ETH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/A6G7TRL7SKK57KCSH27POC4ETH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/A6G7TRL7SKK57KCSH27POC4ETH/action/storage_attestation","attest_author":"https://pith.science/pith/A6G7TRL7SKK57KCSH27POC4ETH/action/author_attestation","sign_citation":"https://pith.science/pith/A6G7TRL7SKK57KCSH27POC4ETH/action/citation_signature","submit_replication":"https://pith.science/pith/A6G7TRL7SKK57KCSH27POC4ETH/action/replication_record"}},"created_at":"2026-06-01T01:03:46.561925+00:00","updated_at":"2026-06-01T01:03:46.561925+00:00"}