{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:Q3PWNCHK42TM77DZP46WUH6U4T","short_pith_number":"pith:Q3PWNCHK","schema_version":"1.0","canonical_sha256":"86df6688eae6a6cffc797f3d6a1fd4e4e7fd5d7849365f121303d9be8e01b160","source":{"kind":"arxiv","id":"2502.06788","version":2},"attestation_state":"computed","paper":{"title":"EVEv2: Improved Baselines for Encoder-Free Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Haiwen Diao, Haoge Deng, Huchuan Lu, Ting Pan, Wenxuan Wang, Xiaotong Li, Xinlong Wang, Yueze Wang, Yufeng Cui","submitted_at":"2025-02-10T18:59:58Z","abstract_excerpt":"Existing encoder-free vision-language models (VLMs) are rapidly narrowing the performance gap with their encoder-based counterparts, highlighting the promising potential for unified multimodal systems with structural simplicity and efficient deployment. We systematically clarify the performance gap between VLMs using pre-trained vision encoders, discrete tokenizers, and minimalist visual layers from scratch, deeply excavating the under-examined characteristics of encoder-free VLMs. We develop efficient strategies for encoder-free VLMs that rival mainstream encoder-based ones. After an in-depth"},"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":"2502.06788","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-02-10T18:59:58Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"85a38fb4801ae555617b4b8237dc79632b269c0039831d7415f5f61f879338cf","abstract_canon_sha256":"0f17897ce7da2b85278a6c9e0f99401ebb130dd9625d090a50b5a9b7c5d70597"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:42:25.793520Z","signature_b64":"fxhXKhqwhNrpfffkyzG9iSBd8jnGBeSyHixCipabN4wGwb+L9lYD/G+Ayj7DlRnVHOplGyFrBtcyRxdBgwq9Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"86df6688eae6a6cffc797f3d6a1fd4e4e7fd5d7849365f121303d9be8e01b160","last_reissued_at":"2026-07-05T11:42:25.792942Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:42:25.792942Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"EVEv2: Improved Baselines for Encoder-Free Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Haiwen Diao, Haoge Deng, Huchuan Lu, Ting Pan, Wenxuan Wang, Xiaotong Li, Xinlong Wang, Yueze Wang, Yufeng Cui","submitted_at":"2025-02-10T18:59:58Z","abstract_excerpt":"Existing encoder-free vision-language models (VLMs) are rapidly narrowing the performance gap with their encoder-based counterparts, highlighting the promising potential for unified multimodal systems with structural simplicity and efficient deployment. We systematically clarify the performance gap between VLMs using pre-trained vision encoders, discrete tokenizers, and minimalist visual layers from scratch, deeply excavating the under-examined characteristics of encoder-free VLMs. We develop efficient strategies for encoder-free VLMs that rival mainstream encoder-based ones. After an in-depth"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06788","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/2502.06788/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":"2502.06788","created_at":"2026-07-05T11:42:25.793011+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.06788v2","created_at":"2026-07-05T11:42:25.793011+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06788","created_at":"2026-07-05T11:42:25.793011+00:00"},{"alias_kind":"pith_short_12","alias_value":"Q3PWNCHK42TM","created_at":"2026-07-05T11:42:25.793011+00:00"},{"alias_kind":"pith_short_16","alias_value":"Q3PWNCHK42TM77DZ","created_at":"2026-07-05T11:42:25.793011+00:00"},{"alias_kind":"pith_short_8","alias_value":"Q3PWNCHK","created_at":"2026-07-05T11:42:25.793011+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.15564","citing_title":"Show-o2: Improved Native Unified Multimodal Models","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2506.01844","citing_title":"SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Q3PWNCHK42TM77DZP46WUH6U4T","json":"https://pith.science/pith/Q3PWNCHK42TM77DZP46WUH6U4T.json","graph_json":"https://pith.science/api/pith-number/Q3PWNCHK42TM77DZP46WUH6U4T/graph.json","events_json":"https://pith.science/api/pith-number/Q3PWNCHK42TM77DZP46WUH6U4T/events.json","paper":"https://pith.science/paper/Q3PWNCHK"},"agent_actions":{"view_html":"https://pith.science/pith/Q3PWNCHK42TM77DZP46WUH6U4T","download_json":"https://pith.science/pith/Q3PWNCHK42TM77DZP46WUH6U4T.json","view_paper":"https://pith.science/paper/Q3PWNCHK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.06788&json=true","fetch_graph":"https://pith.science/api/pith-number/Q3PWNCHK42TM77DZP46WUH6U4T/graph.json","fetch_events":"https://pith.science/api/pith-number/Q3PWNCHK42TM77DZP46WUH6U4T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Q3PWNCHK42TM77DZP46WUH6U4T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Q3PWNCHK42TM77DZP46WUH6U4T/action/storage_attestation","attest_author":"https://pith.science/pith/Q3PWNCHK42TM77DZP46WUH6U4T/action/author_attestation","sign_citation":"https://pith.science/pith/Q3PWNCHK42TM77DZP46WUH6U4T/action/citation_signature","submit_replication":"https://pith.science/pith/Q3PWNCHK42TM77DZP46WUH6U4T/action/replication_record"}},"created_at":"2026-07-05T11:42:25.793011+00:00","updated_at":"2026-07-05T11:42:25.793011+00:00"}