{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3XCKCX75YHLNQEDAYBBQVC257V","short_pith_number":"pith:3XCKCX75","schema_version":"1.0","canonical_sha256":"ddc4a15ffdc1d6d81060c0430a8b5dfd51e1390c5a5d3764c8717067bfb4f021","source":{"kind":"arxiv","id":"2605.10239","version":2},"attestation_state":"computed","paper":{"title":"AdaptSplat: Adapting Vision Foundation Models for Feed-Forward 3D Gaussian Splatting","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"A single 1.5-million-parameter adapter added to vision foundation models enables superior feed-forward 3D Gaussian Splatting.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Mingwei Xing, Xinliang Wang, Yifeng Shi","submitted_at":"2026-05-11T09:14:05Z","abstract_excerpt":"This work explores a simple yet powerful lightweight adapter design for feed-forward 3D Gaussian Splatting (3DGS). Existing methods typically apply complex, architecture-specific designs on top of the generic pipeline of image feature extraction $\\rightarrow$ multi-view interaction $\\rightarrow$ feature decoding. However, constrained by the scale bottleneck of 3D training data and the low-pass filtering effect of deep networks, these methods still fall short in cross-domain generalization and high-frequency geometric fidelity. To address these problems, we propose AdaptSplat, which demonstrate"},"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":true},"canonical_record":{"source":{"id":"2605.10239","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-11T09:14:05Z","cross_cats_sorted":[],"title_canon_sha256":"4919499c3f803d4a011e7a86b24d5f340a0b04254d5e5ba8747c788cf0127e1c","abstract_canon_sha256":"c6e6280b3afd06a7bfc97586dca381463660936d93e2ff124c2a005c7bf0e8c3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:04:35.719046Z","signature_b64":"HqwS43GnbMlutqtqdEWc2ufe0DvNSU3K0dWF2MkjCHWVqP/59fiz8d3PbgnsHqTVMBNohe5jHxDJGuX65dBcCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ddc4a15ffdc1d6d81060c0430a8b5dfd51e1390c5a5d3764c8717067bfb4f021","last_reissued_at":"2026-05-20T00:04:35.718146Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:04:35.718146Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AdaptSplat: Adapting Vision Foundation Models for Feed-Forward 3D Gaussian Splatting","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"A single 1.5-million-parameter adapter added to vision foundation models enables superior feed-forward 3D Gaussian Splatting.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Mingwei Xing, Xinliang Wang, Yifeng Shi","submitted_at":"2026-05-11T09:14:05Z","abstract_excerpt":"This work explores a simple yet powerful lightweight adapter design for feed-forward 3D Gaussian Splatting (3DGS). Existing methods typically apply complex, architecture-specific designs on top of the generic pipeline of image feature extraction $\\rightarrow$ multi-view interaction $\\rightarrow$ feature decoding. However, constrained by the scale bottleneck of 3D training data and the low-pass filtering effect of deep networks, these methods still fall short in cross-domain generalization and high-frequency geometric fidelity. To address these problems, we propose AdaptSplat, which demonstrate"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"introducing a single adapter of only 1.5M parameters into the generic architecture is sufficient to achieve superior performance.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That high-frequency structural priors extracted from shallow features of a vision foundation model backbone can be seamlessly integrated to compensate for low-pass filtering without requiring additional domain-specific fine-tuning or larger training data.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"AdaptSplat adds a lightweight Frequency-Preserving Adapter to vision foundation models that extracts direction-aware high-frequency priors and integrates them via positional encodings and residual modulation to improve feed-forward 3D Gaussian Splatting fidelity and cross-domain performance.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A single 1.5-million-parameter adapter added to vision foundation models enables superior feed-forward 3D Gaussian Splatting.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"348351d973d84d6400925f37a7e9e5609f38d2d0a5204f2abf720f68d64c6cb8"},"source":{"id":"2605.10239","kind":"arxiv","version":2},"verdict":{"id":"8c551969-000e-4cf4-a58b-7df4ea1dbbc9","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-12T03:25:46.238373Z","strongest_claim":"introducing a single adapter of only 1.5M parameters into the generic architecture is sufficient to achieve superior performance.","one_line_summary":"AdaptSplat adds a lightweight Frequency-Preserving Adapter to vision foundation models that extracts direction-aware high-frequency priors and integrates them via positional encodings and residual modulation to improve feed-forward 3D Gaussian Splatting fidelity and cross-domain performance.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That high-frequency structural priors extracted from shallow features of a vision foundation model backbone can be seamlessly integrated to compensate for low-pass filtering without requiring additional domain-specific fine-tuning or larger training data.","pith_extraction_headline":"A single 1.5-million-parameter adapter added to vision foundation models enables superior feed-forward 3D Gaussian Splatting."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.10239/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-19T15:38:03.960704Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T11:31:19.712732Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T09:33:05.558837Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"40fd28a32eb7a970c5ebf75602105a00cd1a317ebfbd9154fd03898a50ad6555"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"69d6a262322efcbca748cfad2839f293b00834daccc562f6830a5271ece9d9d8"},"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":"2605.10239","created_at":"2026-05-20T00:04:35.718329+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.10239v2","created_at":"2026-05-20T00:04:35.718329+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.10239","created_at":"2026-05-20T00:04:35.718329+00:00"},{"alias_kind":"pith_short_12","alias_value":"3XCKCX75YHLN","created_at":"2026-05-20T00:04:35.718329+00:00"},{"alias_kind":"pith_short_16","alias_value":"3XCKCX75YHLNQEDA","created_at":"2026-05-20T00:04:35.718329+00:00"},{"alias_kind":"pith_short_8","alias_value":"3XCKCX75","created_at":"2026-05-20T00:04:35.718329+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3XCKCX75YHLNQEDAYBBQVC257V","json":"https://pith.science/pith/3XCKCX75YHLNQEDAYBBQVC257V.json","graph_json":"https://pith.science/api/pith-number/3XCKCX75YHLNQEDAYBBQVC257V/graph.json","events_json":"https://pith.science/api/pith-number/3XCKCX75YHLNQEDAYBBQVC257V/events.json","paper":"https://pith.science/paper/3XCKCX75"},"agent_actions":{"view_html":"https://pith.science/pith/3XCKCX75YHLNQEDAYBBQVC257V","download_json":"https://pith.science/pith/3XCKCX75YHLNQEDAYBBQVC257V.json","view_paper":"https://pith.science/paper/3XCKCX75","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.10239&json=true","fetch_graph":"https://pith.science/api/pith-number/3XCKCX75YHLNQEDAYBBQVC257V/graph.json","fetch_events":"https://pith.science/api/pith-number/3XCKCX75YHLNQEDAYBBQVC257V/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3XCKCX75YHLNQEDAYBBQVC257V/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3XCKCX75YHLNQEDAYBBQVC257V/action/storage_attestation","attest_author":"https://pith.science/pith/3XCKCX75YHLNQEDAYBBQVC257V/action/author_attestation","sign_citation":"https://pith.science/pith/3XCKCX75YHLNQEDAYBBQVC257V/action/citation_signature","submit_replication":"https://pith.science/pith/3XCKCX75YHLNQEDAYBBQVC257V/action/replication_record"}},"created_at":"2026-05-20T00:04:35.718329+00:00","updated_at":"2026-05-20T00:04:35.718329+00:00"}