{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:JUNX4FSD6BRBCAJ5DWIDALXATB","short_pith_number":"pith:JUNX4FSD","schema_version":"1.0","canonical_sha256":"4d1b7e1643f06211013d1d90302ee0986e6792b944164d6a8e47f846067ab726","source":{"kind":"arxiv","id":"2410.13461","version":2},"attestation_state":"computed","paper":{"title":"Progressive Mixed-Precision Decoding for Efficient LLM Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Alexandros Kouris, Fuwen Tan, Hao Mark Chen, Hongxiang Fan, Royson Lee, Stylianos I. Venieris","submitted_at":"2024-10-17T11:46:33Z","abstract_excerpt":"In spite of the great potential of large language models (LLMs) across various tasks, their deployment on resource-constrained devices remains challenging due to their excessive computational and memory demands. Quantization has emerged as an effective solution by storing weights in reduced precision. However, utilizing low precisions (i.e.~2/3-bit) to substantially alleviate the memory-boundedness of LLM decoding, still suffers from prohibitive performance drop. In this work, we argue that existing approaches fail to explore the diversity in computational patterns, redundancy, and sensitivity"},"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":"2410.13461","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-17T11:46:33Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"def5b49a48657abdf5d69ce3c94df5f2f608578cf254f26b4930d09ba62b577c","abstract_canon_sha256":"10a60cb92ead480a3c8c7ec1b9afabe1cacb065e60837a049ad166c46f0d3ec4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:20:40.828305Z","signature_b64":"50i83gI81KC/lSM0cOCBFKg0FIM4rbn4kyp2Hr/QfKpLl4aVH1anTPQnrf+JtOoEu6/Ih8UU3ZDigz+BcHu2Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4d1b7e1643f06211013d1d90302ee0986e6792b944164d6a8e47f846067ab726","last_reissued_at":"2026-07-05T10:20:40.827211Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:20:40.827211Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Progressive Mixed-Precision Decoding for Efficient LLM Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Alexandros Kouris, Fuwen Tan, Hao Mark Chen, Hongxiang Fan, Royson Lee, Stylianos I. Venieris","submitted_at":"2024-10-17T11:46:33Z","abstract_excerpt":"In spite of the great potential of large language models (LLMs) across various tasks, their deployment on resource-constrained devices remains challenging due to their excessive computational and memory demands. Quantization has emerged as an effective solution by storing weights in reduced precision. However, utilizing low precisions (i.e.~2/3-bit) to substantially alleviate the memory-boundedness of LLM decoding, still suffers from prohibitive performance drop. In this work, we argue that existing approaches fail to explore the diversity in computational patterns, redundancy, and sensitivity"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.13461","kind":"arxiv","version":2},"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/2410.13461/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":"2410.13461","created_at":"2026-07-05T10:20:40.827309+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.13461v2","created_at":"2026-07-05T10:20:40.827309+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.13461","created_at":"2026-07-05T10:20:40.827309+00:00"},{"alias_kind":"pith_short_12","alias_value":"JUNX4FSD6BRB","created_at":"2026-07-05T10:20:40.827309+00:00"},{"alias_kind":"pith_short_16","alias_value":"JUNX4FSD6BRBCAJ5","created_at":"2026-07-05T10:20:40.827309+00:00"},{"alias_kind":"pith_short_8","alias_value":"JUNX4FSD","created_at":"2026-07-05T10:20:40.827309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.23419","citing_title":"GRINQH: Graded Input-based Quantization Hierarchy for Efficient LLM Generation","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29708","citing_title":"Demystifying the Design Space and Best Practices for Heterogeneous LLM Inference and Serving","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29708","citing_title":"Demystifying the Design Space and Best Practices for Heterogeneous LLM Inference and Serving","ref_index":25,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JUNX4FSD6BRBCAJ5DWIDALXATB","json":"https://pith.science/pith/JUNX4FSD6BRBCAJ5DWIDALXATB.json","graph_json":"https://pith.science/api/pith-number/JUNX4FSD6BRBCAJ5DWIDALXATB/graph.json","events_json":"https://pith.science/api/pith-number/JUNX4FSD6BRBCAJ5DWIDALXATB/events.json","paper":"https://pith.science/paper/JUNX4FSD"},"agent_actions":{"view_html":"https://pith.science/pith/JUNX4FSD6BRBCAJ5DWIDALXATB","download_json":"https://pith.science/pith/JUNX4FSD6BRBCAJ5DWIDALXATB.json","view_paper":"https://pith.science/paper/JUNX4FSD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.13461&json=true","fetch_graph":"https://pith.science/api/pith-number/JUNX4FSD6BRBCAJ5DWIDALXATB/graph.json","fetch_events":"https://pith.science/api/pith-number/JUNX4FSD6BRBCAJ5DWIDALXATB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JUNX4FSD6BRBCAJ5DWIDALXATB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JUNX4FSD6BRBCAJ5DWIDALXATB/action/storage_attestation","attest_author":"https://pith.science/pith/JUNX4FSD6BRBCAJ5DWIDALXATB/action/author_attestation","sign_citation":"https://pith.science/pith/JUNX4FSD6BRBCAJ5DWIDALXATB/action/citation_signature","submit_replication":"https://pith.science/pith/JUNX4FSD6BRBCAJ5DWIDALXATB/action/replication_record"}},"created_at":"2026-07-05T10:20:40.827309+00:00","updated_at":"2026-07-05T10:20:40.827309+00:00"}