{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:22IF2DVMECP6TRMNDRTURVC4SC","short_pith_number":"pith:22IF2DVM","schema_version":"1.0","canonical_sha256":"d6905d0eac209fe9c58d1c6748d45c908b1e00f1213b5e0b4283055dd5e82565","source":{"kind":"arxiv","id":"2508.12590","version":1},"attestation_state":"computed","paper":{"title":"Energy-Efficient Wireless LLM Inference via Uncertainty and Importance-Aware Speculative Decoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jihoon Park, Seong-Lyun Kim, Seungeun Oh","submitted_at":"2025-08-18T02:56:59Z","abstract_excerpt":"To address the growing demand for on-device LLM inference in resource-constrained environments, hybrid language models (HLM) have emerged, combining lightweight local models with powerful cloud-based LLMs. Recent studies on HLM have primarily focused on improving accuracy and latency, while often overlooking communication and energy efficiency. We propose a token-level filtering mechanism for an energy-efficient importance- and uncertainty-aware HLM inference that leverages both epistemic uncertainty and attention-based importance. Our method opportunistically uploads only informative tokens, "},"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":"2508.12590","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-18T02:56:59Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9d6d50377125783bd62a1664091090070e03efb77e60cc00596c43fb726a3274","abstract_canon_sha256":"ae6ec5bdbccdfa040f06a5c145d1780bf37d17d67bce0e92d0fc67614d01f396"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:55:23.637186Z","signature_b64":"usVjGoA8OzsJqVwsY106kblDAlEnsPXbWdI6DEtabhcOKTUmacbILNC9u3OOEqExZ6F0OQTUV0fwlUH3eKSVCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d6905d0eac209fe9c58d1c6748d45c908b1e00f1213b5e0b4283055dd5e82565","last_reissued_at":"2026-07-05T11:55:23.636805Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:55:23.636805Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Energy-Efficient Wireless LLM Inference via Uncertainty and Importance-Aware Speculative Decoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jihoon Park, Seong-Lyun Kim, Seungeun Oh","submitted_at":"2025-08-18T02:56:59Z","abstract_excerpt":"To address the growing demand for on-device LLM inference in resource-constrained environments, hybrid language models (HLM) have emerged, combining lightweight local models with powerful cloud-based LLMs. Recent studies on HLM have primarily focused on improving accuracy and latency, while often overlooking communication and energy efficiency. We propose a token-level filtering mechanism for an energy-efficient importance- and uncertainty-aware HLM inference that leverages both epistemic uncertainty and attention-based importance. Our method opportunistically uploads only informative tokens, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.12590","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/2508.12590/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":"2508.12590","created_at":"2026-07-05T11:55:23.636863+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.12590v1","created_at":"2026-07-05T11:55:23.636863+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.12590","created_at":"2026-07-05T11:55:23.636863+00:00"},{"alias_kind":"pith_short_12","alias_value":"22IF2DVMECP6","created_at":"2026-07-05T11:55:23.636863+00:00"},{"alias_kind":"pith_short_16","alias_value":"22IF2DVMECP6TRMN","created_at":"2026-07-05T11:55:23.636863+00:00"},{"alias_kind":"pith_short_8","alias_value":"22IF2DVM","created_at":"2026-07-05T11:55:23.636863+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/22IF2DVMECP6TRMNDRTURVC4SC","json":"https://pith.science/pith/22IF2DVMECP6TRMNDRTURVC4SC.json","graph_json":"https://pith.science/api/pith-number/22IF2DVMECP6TRMNDRTURVC4SC/graph.json","events_json":"https://pith.science/api/pith-number/22IF2DVMECP6TRMNDRTURVC4SC/events.json","paper":"https://pith.science/paper/22IF2DVM"},"agent_actions":{"view_html":"https://pith.science/pith/22IF2DVMECP6TRMNDRTURVC4SC","download_json":"https://pith.science/pith/22IF2DVMECP6TRMNDRTURVC4SC.json","view_paper":"https://pith.science/paper/22IF2DVM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.12590&json=true","fetch_graph":"https://pith.science/api/pith-number/22IF2DVMECP6TRMNDRTURVC4SC/graph.json","fetch_events":"https://pith.science/api/pith-number/22IF2DVMECP6TRMNDRTURVC4SC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/22IF2DVMECP6TRMNDRTURVC4SC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/22IF2DVMECP6TRMNDRTURVC4SC/action/storage_attestation","attest_author":"https://pith.science/pith/22IF2DVMECP6TRMNDRTURVC4SC/action/author_attestation","sign_citation":"https://pith.science/pith/22IF2DVMECP6TRMNDRTURVC4SC/action/citation_signature","submit_replication":"https://pith.science/pith/22IF2DVMECP6TRMNDRTURVC4SC/action/replication_record"}},"created_at":"2026-07-05T11:55:23.636863+00:00","updated_at":"2026-07-05T11:55:23.636863+00:00"}