{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4BS7OEQGDASJXS2XHDTAFHENGI","short_pith_number":"pith:4BS7OEQG","schema_version":"1.0","canonical_sha256":"e065f7120618249bcb5738e6029c8d3232c41db59de0bc7205c820e47f6d18fb","source":{"kind":"arxiv","id":"2606.03879","version":1},"attestation_state":"computed","paper":{"title":"Beyond Encoder Accumulation: Measuring Encoder Roles in Multi-Encoder VLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Jiansheng Chen, Ruobing Xie, Wei Ding, Xingwu Sun, Yudong Zhang, Yu Wang","submitted_at":"2026-06-02T16:46:42Z","abstract_excerpt":"As foundation models scale toward fusing more heterogeneous visual streams, understanding how diverse encoders interact under joint training becomes a prerequisite for principled design. Yet large vision-language models (LVLMs) currently lack the tools to do so, and parameter-efficient encoder configurations remain hard to identify before training. To re-examine encoder roles under joint training, on the 16-benchmark Cambrian-1 suite we retrain and evaluate all 31 non-empty subsets of five common vision encoders under a unified pipeline (~20k GPU-hours total), and report three findings. First,"},"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":"2606.03879","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-06-02T16:46:42Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c0f24677b0bd55b8e61177673644b50a4e9d117b09eed80668c09c52f0527ed3","abstract_canon_sha256":"11f9110b8d2393f002676d8cf3d7468c4b6847a31380ccccdc42d70c4c9faa53"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-03T02:06:05.126491Z","signature_b64":"Lj3++h4oWKCBf5hwEHglTO4CerYitcy0nWmO+/lSdEsiREWgd8JAlUzinXVAy1pAnyoEk65KWDozyM5Fq1mjBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e065f7120618249bcb5738e6029c8d3232c41db59de0bc7205c820e47f6d18fb","last_reissued_at":"2026-06-03T02:06:05.126126Z","signature_status":"signed_v1","first_computed_at":"2026-06-03T02:06:05.126126Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Encoder Accumulation: Measuring Encoder Roles in Multi-Encoder VLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Jiansheng Chen, Ruobing Xie, Wei Ding, Xingwu Sun, Yudong Zhang, Yu Wang","submitted_at":"2026-06-02T16:46:42Z","abstract_excerpt":"As foundation models scale toward fusing more heterogeneous visual streams, understanding how diverse encoders interact under joint training becomes a prerequisite for principled design. Yet large vision-language models (LVLMs) currently lack the tools to do so, and parameter-efficient encoder configurations remain hard to identify before training. To re-examine encoder roles under joint training, on the 16-benchmark Cambrian-1 suite we retrain and evaluate all 31 non-empty subsets of five common vision encoders under a unified pipeline (~20k GPU-hours total), and report three findings. First,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.03879","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/2606.03879/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":"2606.03879","created_at":"2026-06-03T02:06:05.126182+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.03879v1","created_at":"2026-06-03T02:06:05.126182+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.03879","created_at":"2026-06-03T02:06:05.126182+00:00"},{"alias_kind":"pith_short_12","alias_value":"4BS7OEQGDASJ","created_at":"2026-06-03T02:06:05.126182+00:00"},{"alias_kind":"pith_short_16","alias_value":"4BS7OEQGDASJXS2X","created_at":"2026-06-03T02:06:05.126182+00:00"},{"alias_kind":"pith_short_8","alias_value":"4BS7OEQG","created_at":"2026-06-03T02:06:05.126182+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/4BS7OEQGDASJXS2XHDTAFHENGI","json":"https://pith.science/pith/4BS7OEQGDASJXS2XHDTAFHENGI.json","graph_json":"https://pith.science/api/pith-number/4BS7OEQGDASJXS2XHDTAFHENGI/graph.json","events_json":"https://pith.science/api/pith-number/4BS7OEQGDASJXS2XHDTAFHENGI/events.json","paper":"https://pith.science/paper/4BS7OEQG"},"agent_actions":{"view_html":"https://pith.science/pith/4BS7OEQGDASJXS2XHDTAFHENGI","download_json":"https://pith.science/pith/4BS7OEQGDASJXS2XHDTAFHENGI.json","view_paper":"https://pith.science/paper/4BS7OEQG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.03879&json=true","fetch_graph":"https://pith.science/api/pith-number/4BS7OEQGDASJXS2XHDTAFHENGI/graph.json","fetch_events":"https://pith.science/api/pith-number/4BS7OEQGDASJXS2XHDTAFHENGI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4BS7OEQGDASJXS2XHDTAFHENGI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4BS7OEQGDASJXS2XHDTAFHENGI/action/storage_attestation","attest_author":"https://pith.science/pith/4BS7OEQGDASJXS2XHDTAFHENGI/action/author_attestation","sign_citation":"https://pith.science/pith/4BS7OEQGDASJXS2XHDTAFHENGI/action/citation_signature","submit_replication":"https://pith.science/pith/4BS7OEQGDASJXS2XHDTAFHENGI/action/replication_record"}},"created_at":"2026-06-03T02:06:05.126182+00:00","updated_at":"2026-06-03T02:06:05.126182+00:00"}