{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RZJ6CRDZQOR62ELFWMVIF5EPGG","short_pith_number":"pith:RZJ6CRDZ","schema_version":"1.0","canonical_sha256":"8e53e1447983a3ed1165b32a82f48f31a17f7aee904b8d3628ce6b5a8c244188","source":{"kind":"arxiv","id":"2508.02128","version":1},"attestation_state":"computed","paper":{"title":"Amber Pruner: Leveraging N:M Activation Sparsity for Efficient Prefill in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Gongyi Wang, Hao Chen, Pengcheng Xie, Ruwu Cai, Sheng Chang, Tai An, Yang Liu, Yanzhe Zhang, Yiwu Yao","submitted_at":"2025-08-04T07:22:36Z","abstract_excerpt":"In the era of large language models (LLMs), N:M sparsity has emerged as a structured compression technique critical for accelerating inference. While prior work has primarily focused on weight sparsity, it often suffers from significant accuracy degradation. Activation sparsity, though promising, is typically training-dependent and faces challenges in generalization. To address these limitations, we introduce Amber Pruner, a training-free N:M activation sparsity method designed specifically for the prefill stage, targeting the acceleration of linear projection layers in LLMs. Extensive experim"},"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":"2508.02128","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-04T07:22:36Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"19ee75c147a2f655993286a9ccf974abbf6261c5edc006cb14059e60976a7546","abstract_canon_sha256":"831ae7ccb242e20297f8f32d6509c2d24ef75fe4e425880c62acc592406007ac"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:48:08.688748Z","signature_b64":"JylTa9Mo+8Ub8A1KNKD53DrKXchRaWAccQeTOFM7FSX2oWBPMmLwcWveFH4d8UsRxdDerfngthiB5cQYfnGlDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8e53e1447983a3ed1165b32a82f48f31a17f7aee904b8d3628ce6b5a8c244188","last_reissued_at":"2026-07-05T11:48:08.688236Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:48:08.688236Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Amber Pruner: Leveraging N:M Activation Sparsity for Efficient Prefill in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Gongyi Wang, Hao Chen, Pengcheng Xie, Ruwu Cai, Sheng Chang, Tai An, Yang Liu, Yanzhe Zhang, Yiwu Yao","submitted_at":"2025-08-04T07:22:36Z","abstract_excerpt":"In the era of large language models (LLMs), N:M sparsity has emerged as a structured compression technique critical for accelerating inference. While prior work has primarily focused on weight sparsity, it often suffers from significant accuracy degradation. Activation sparsity, though promising, is typically training-dependent and faces challenges in generalization. To address these limitations, we introduce Amber Pruner, a training-free N:M activation sparsity method designed specifically for the prefill stage, targeting the acceleration of linear projection layers in LLMs. Extensive experim"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.02128","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/2508.02128/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":"2508.02128","created_at":"2026-07-05T11:48:08.688299+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.02128v1","created_at":"2026-07-05T11:48:08.688299+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.02128","created_at":"2026-07-05T11:48:08.688299+00:00"},{"alias_kind":"pith_short_12","alias_value":"RZJ6CRDZQOR6","created_at":"2026-07-05T11:48:08.688299+00:00"},{"alias_kind":"pith_short_16","alias_value":"RZJ6CRDZQOR62ELF","created_at":"2026-07-05T11:48:08.688299+00:00"},{"alias_kind":"pith_short_8","alias_value":"RZJ6CRDZ","created_at":"2026-07-05T11:48:08.688299+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.26632","citing_title":"RT-Lynx: Putting GEMM Sparsity in the Right Place for Diffusion Models","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2509.22166","citing_title":"Motivating Next-Gen Accelerators with Flexible (N:M) Activation Sparsity via Benchmarking Lightweight Post-Training Sparsification Approaches","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RZJ6CRDZQOR62ELFWMVIF5EPGG","json":"https://pith.science/pith/RZJ6CRDZQOR62ELFWMVIF5EPGG.json","graph_json":"https://pith.science/api/pith-number/RZJ6CRDZQOR62ELFWMVIF5EPGG/graph.json","events_json":"https://pith.science/api/pith-number/RZJ6CRDZQOR62ELFWMVIF5EPGG/events.json","paper":"https://pith.science/paper/RZJ6CRDZ"},"agent_actions":{"view_html":"https://pith.science/pith/RZJ6CRDZQOR62ELFWMVIF5EPGG","download_json":"https://pith.science/pith/RZJ6CRDZQOR62ELFWMVIF5EPGG.json","view_paper":"https://pith.science/paper/RZJ6CRDZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.02128&json=true","fetch_graph":"https://pith.science/api/pith-number/RZJ6CRDZQOR62ELFWMVIF5EPGG/graph.json","fetch_events":"https://pith.science/api/pith-number/RZJ6CRDZQOR62ELFWMVIF5EPGG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RZJ6CRDZQOR62ELFWMVIF5EPGG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RZJ6CRDZQOR62ELFWMVIF5EPGG/action/storage_attestation","attest_author":"https://pith.science/pith/RZJ6CRDZQOR62ELFWMVIF5EPGG/action/author_attestation","sign_citation":"https://pith.science/pith/RZJ6CRDZQOR62ELFWMVIF5EPGG/action/citation_signature","submit_replication":"https://pith.science/pith/RZJ6CRDZQOR62ELFWMVIF5EPGG/action/replication_record"}},"created_at":"2026-07-05T11:48:08.688299+00:00","updated_at":"2026-07-05T11:48:08.688299+00:00"}