{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:BUTFEGRN4GJ4DEDYIKJXHI43NZ","short_pith_number":"pith:BUTFEGRN","schema_version":"1.0","canonical_sha256":"0d26521a2de193c19078429373a39b6e53c6e23ca0e57531ea4ab6421c44d960","source":{"kind":"arxiv","id":"2606.08094","version":1},"attestation_state":"computed","paper":{"title":"vla.cpp: A Unified Inference Runtime for Vision-Language-Action Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.SY","eess.SY"],"primary_cat":"cs.RO","authors_text":"An T. Le, Chinh T. Nguyen, Duy M. H. Nguyen, Hung T. Ho, Khanh D. Nguyen, Linh D. Le, Thanh Q. Duong, Vien A. Ngo","submitted_at":"2026-06-06T10:45:40Z","abstract_excerpt":"Vision-Language-Action (VLA) policies are typically shipped as Python/PyTorch stacks that assume a workstation-class GPU, a mismatch for the hardware on which robots actually run. We present vla.cpp, a portable C++ inference runtime built on llama.cpp. To our knowledge, it is the first ggml-class engine to natively serve the flow-matching and diffusion VLA inference pattern, in which a cached vision-language prefix is consumed by a cross-attending action expert integrated over several solver steps. A single runtime serves seven architectures spanning five backbone and four action-head families"},"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":"2606.08094","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-06-06T10:45:40Z","cross_cats_sorted":["cs.AI","cs.LG","cs.SY","eess.SY"],"title_canon_sha256":"583cb9adc5d1678eb7e0b00414caaa958050abd7c0051ec000f8fa574d45118a","abstract_canon_sha256":"012a792db567f0aaa1a849ee804c44dab9dd399a8e3ed9d063e299e14b5c1f20"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T01:05:26.224498Z","signature_b64":"6K87rw/vTZDUGUyIb655X/15TlFoH8D//LIBqAC9KQ5KzclOy6Fo2OG3iHsv5RM3ZuQRVlUgh65O3VfWlPIZDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0d26521a2de193c19078429373a39b6e53c6e23ca0e57531ea4ab6421c44d960","last_reissued_at":"2026-06-09T01:05:26.224035Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T01:05:26.224035Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"vla.cpp: A Unified Inference Runtime for Vision-Language-Action Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.SY","eess.SY"],"primary_cat":"cs.RO","authors_text":"An T. Le, Chinh T. Nguyen, Duy M. H. Nguyen, Hung T. Ho, Khanh D. Nguyen, Linh D. Le, Thanh Q. Duong, Vien A. Ngo","submitted_at":"2026-06-06T10:45:40Z","abstract_excerpt":"Vision-Language-Action (VLA) policies are typically shipped as Python/PyTorch stacks that assume a workstation-class GPU, a mismatch for the hardware on which robots actually run. We present vla.cpp, a portable C++ inference runtime built on llama.cpp. To our knowledge, it is the first ggml-class engine to natively serve the flow-matching and diffusion VLA inference pattern, in which a cached vision-language prefix is consumed by a cross-attending action expert integrated over several solver steps. A single runtime serves seven architectures spanning five backbone and four action-head families"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.08094","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/2606.08094/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":"2606.08094","created_at":"2026-06-09T01:05:26.224102+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.08094v1","created_at":"2026-06-09T01:05:26.224102+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.08094","created_at":"2026-06-09T01:05:26.224102+00:00"},{"alias_kind":"pith_short_12","alias_value":"BUTFEGRN4GJ4","created_at":"2026-06-09T01:05:26.224102+00:00"},{"alias_kind":"pith_short_16","alias_value":"BUTFEGRN4GJ4DEDY","created_at":"2026-06-09T01:05:26.224102+00:00"},{"alias_kind":"pith_short_8","alias_value":"BUTFEGRN","created_at":"2026-06-09T01:05:26.224102+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/BUTFEGRN4GJ4DEDYIKJXHI43NZ","json":"https://pith.science/pith/BUTFEGRN4GJ4DEDYIKJXHI43NZ.json","graph_json":"https://pith.science/api/pith-number/BUTFEGRN4GJ4DEDYIKJXHI43NZ/graph.json","events_json":"https://pith.science/api/pith-number/BUTFEGRN4GJ4DEDYIKJXHI43NZ/events.json","paper":"https://pith.science/paper/BUTFEGRN"},"agent_actions":{"view_html":"https://pith.science/pith/BUTFEGRN4GJ4DEDYIKJXHI43NZ","download_json":"https://pith.science/pith/BUTFEGRN4GJ4DEDYIKJXHI43NZ.json","view_paper":"https://pith.science/paper/BUTFEGRN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.08094&json=true","fetch_graph":"https://pith.science/api/pith-number/BUTFEGRN4GJ4DEDYIKJXHI43NZ/graph.json","fetch_events":"https://pith.science/api/pith-number/BUTFEGRN4GJ4DEDYIKJXHI43NZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BUTFEGRN4GJ4DEDYIKJXHI43NZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BUTFEGRN4GJ4DEDYIKJXHI43NZ/action/storage_attestation","attest_author":"https://pith.science/pith/BUTFEGRN4GJ4DEDYIKJXHI43NZ/action/author_attestation","sign_citation":"https://pith.science/pith/BUTFEGRN4GJ4DEDYIKJXHI43NZ/action/citation_signature","submit_replication":"https://pith.science/pith/BUTFEGRN4GJ4DEDYIKJXHI43NZ/action/replication_record"}},"created_at":"2026-06-09T01:05:26.224102+00:00","updated_at":"2026-06-09T01:05:26.224102+00:00"}