{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:64OFD2THLAG3IOR2YYKMT7A2G3","short_pith_number":"pith:64OFD2TH","schema_version":"1.0","canonical_sha256":"f71c51ea67580db43a3ac614c9fc1a36ddf504f1d14ca24274aac5b8e4cddc6f","source":{"kind":"arxiv","id":"2608.04428","version":1},"attestation_state":"computed","paper":{"title":"Deltoris: Enabling Real-time VLA Inference in Embodied AI via Bit-level Sparsity and Speculative Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AR","authors_text":"Anbang Wu, Fangxin Liu, Han Zhao, Jingwen Leng, Minyi Guo, Yiming Gan, Yinhe Han, Yu Feng, Zeyu Guo, Zheng Liu, Zhezhi He, Zihan Liu","submitted_at":"2026-08-05T04:17:03Z","abstract_excerpt":"Vision-language-action (VLA) models have emerged as a key component in embodied AI. Among existing approaches, diffusion-based VLA models achieve superior motion quality and generalization. However, diffusion-based VLA models are compute-intensive and must run at high control frequency, e.g., 50-200 Hz. Thus, it imposes strict latency and energy constraints on edge devices.\n  In this work, we present Deltoris, an algorithm-hardware co-design framework for efficient diffusion-based VLA inference. First, we exploit the temporal similarity of consecutive inputs and propose a \\textit{temporal-awar"},"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":"2608.04428","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AR","submitted_at":"2026-08-05T04:17:03Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"dcc9c30224a357aac6678c32587a3f8a2c6b0cd001dbe3653c766e70e6b05a9e","abstract_canon_sha256":"2e05a120e78d17156ed2db70e701ab1c89c3647863c712a5ea23560a25486b21"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-06T01:35:24.714052Z","signature_b64":"srDPQXv4CrT3XVaevRETqklGNKIdoQLxS0/Y1grf7mwyfGdYmwjC9o9pWJN8KkwRZrcnzdl+GU8gl0f1DIBgBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f71c51ea67580db43a3ac614c9fc1a36ddf504f1d14ca24274aac5b8e4cddc6f","last_reissued_at":"2026-08-06T01:35:24.712563Z","signature_status":"signed_v1","first_computed_at":"2026-08-06T01:35:24.712563Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Deltoris: Enabling Real-time VLA Inference in Embodied AI via Bit-level Sparsity and Speculative Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AR","authors_text":"Anbang Wu, Fangxin Liu, Han Zhao, Jingwen Leng, Minyi Guo, Yiming Gan, Yinhe Han, Yu Feng, Zeyu Guo, Zheng Liu, Zhezhi He, Zihan Liu","submitted_at":"2026-08-05T04:17:03Z","abstract_excerpt":"Vision-language-action (VLA) models have emerged as a key component in embodied AI. Among existing approaches, diffusion-based VLA models achieve superior motion quality and generalization. However, diffusion-based VLA models are compute-intensive and must run at high control frequency, e.g., 50-200 Hz. Thus, it imposes strict latency and energy constraints on edge devices.\n  In this work, we present Deltoris, an algorithm-hardware co-design framework for efficient diffusion-based VLA inference. First, we exploit the temporal similarity of consecutive inputs and propose a \\textit{temporal-awar"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04428","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/2608.04428/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":"2608.04428","created_at":"2026-08-06T01:35:24.714404+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.04428v1","created_at":"2026-08-06T01:35:24.714404+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04428","created_at":"2026-08-06T01:35:24.714404+00:00"},{"alias_kind":"pith_short_12","alias_value":"64OFD2THLAG3","created_at":"2026-08-06T01:35:24.714404+00:00"},{"alias_kind":"pith_short_16","alias_value":"64OFD2THLAG3IOR2","created_at":"2026-08-06T01:35:24.714404+00:00"},{"alias_kind":"pith_short_8","alias_value":"64OFD2TH","created_at":"2026-08-06T01:35:24.714404+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/64OFD2THLAG3IOR2YYKMT7A2G3","json":"https://pith.science/pith/64OFD2THLAG3IOR2YYKMT7A2G3.json","graph_json":"https://pith.science/api/pith-number/64OFD2THLAG3IOR2YYKMT7A2G3/graph.json","events_json":"https://pith.science/api/pith-number/64OFD2THLAG3IOR2YYKMT7A2G3/events.json","paper":"https://pith.science/paper/64OFD2TH"},"agent_actions":{"view_html":"https://pith.science/pith/64OFD2THLAG3IOR2YYKMT7A2G3","download_json":"https://pith.science/pith/64OFD2THLAG3IOR2YYKMT7A2G3.json","view_paper":"https://pith.science/paper/64OFD2TH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.04428&json=true","fetch_graph":"https://pith.science/api/pith-number/64OFD2THLAG3IOR2YYKMT7A2G3/graph.json","fetch_events":"https://pith.science/api/pith-number/64OFD2THLAG3IOR2YYKMT7A2G3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/64OFD2THLAG3IOR2YYKMT7A2G3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/64OFD2THLAG3IOR2YYKMT7A2G3/action/storage_attestation","attest_author":"https://pith.science/pith/64OFD2THLAG3IOR2YYKMT7A2G3/action/author_attestation","sign_citation":"https://pith.science/pith/64OFD2THLAG3IOR2YYKMT7A2G3/action/citation_signature","submit_replication":"https://pith.science/pith/64OFD2THLAG3IOR2YYKMT7A2G3/action/replication_record"}},"created_at":"2026-08-06T01:35:24.714404+00:00","updated_at":"2026-08-06T01:35:24.714404+00:00"}