{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:MJJN7I5Q5KHIAK2DJHDEK4OL2Y","short_pith_number":"pith:MJJN7I5Q","canonical_record":{"source":{"id":"2502.10424","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-05T20:43:48Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1f2e959a2931f5d5a25dc129795dbb0ea2f24fef1056563c44d12f48b33ea13e","abstract_canon_sha256":"59a161c652e7cd125bb52f1a53dd9e2511918b30861cd11d953e38873cbfc7cf"},"schema_version":"1.0"},"canonical_sha256":"6252dfa3b0ea8e802b4349c64571cbd6320333add6751f802b76f17074b8514a","source":{"kind":"arxiv","id":"2502.10424","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.10424","created_at":"2026-07-05T10:14:52Z"},{"alias_kind":"arxiv_version","alias_value":"2502.10424v1","created_at":"2026-07-05T10:14:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.10424","created_at":"2026-07-05T10:14:52Z"},{"alias_kind":"pith_short_12","alias_value":"MJJN7I5Q5KHI","created_at":"2026-07-05T10:14:52Z"},{"alias_kind":"pith_short_16","alias_value":"MJJN7I5Q5KHIAK2D","created_at":"2026-07-05T10:14:52Z"},{"alias_kind":"pith_short_8","alias_value":"MJJN7I5Q","created_at":"2026-07-05T10:14:52Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:MJJN7I5Q5KHIAK2DJHDEK4OL2Y","target":"record","payload":{"canonical_record":{"source":{"id":"2502.10424","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-05T20:43:48Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1f2e959a2931f5d5a25dc129795dbb0ea2f24fef1056563c44d12f48b33ea13e","abstract_canon_sha256":"59a161c652e7cd125bb52f1a53dd9e2511918b30861cd11d953e38873cbfc7cf"},"schema_version":"1.0"},"canonical_sha256":"6252dfa3b0ea8e802b4349c64571cbd6320333add6751f802b76f17074b8514a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:14:52.381129Z","signature_b64":"sWEx/zeyUmeEZ1Vb2F+dJNRpokZukxB9GZxnEiKNd2wrQFymMnYZbTaKFKde1eSSB5m2zMvgUZy0/ENNet+1DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6252dfa3b0ea8e802b4349c64571cbd6320333add6751f802b76f17074b8514a","last_reissued_at":"2026-07-05T10:14:52.380637Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:14:52.380637Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.10424","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:14:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NJ2WDQIy4YkxsyD8xHDv/N+4QiqVYnF3XVeTsN4Q9FlcYSN/q1I9fm4nAsSDZjkvPbeF/bXRNu2kJgU27zVSDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T01:21:36.711684Z"},"content_sha256":"aed677611980c6b5b152a76e018a1214ca60dfcba45f3e5d08538760ed981410","schema_version":"1.0","event_id":"sha256:aed677611980c6b5b152a76e018a1214ca60dfcba45f3e5d08538760ed981410"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:MJJN7I5Q5KHIAK2DJHDEK4OL2Y","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"QuantSpec: Self-Speculative Decoding with Hierarchical Quantized KV Cache","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Aditya Tomar, Amir Gholami, Coleman Hooper, Haocheng Xi, Kurt Keutzer, Mahyar Najibi, Maxwell Horton, Michael W. Mahoney, Rishabh Tiwari, Sehoon Kim","submitted_at":"2025-02-05T20:43:48Z","abstract_excerpt":"Large Language Models (LLMs) are increasingly being deployed on edge devices for long-context settings, creating a growing need for fast and efficient long-context inference. In these scenarios, the Key-Value (KV) cache is the primary bottleneck in terms of both GPU memory and latency, as the full KV cache must be loaded for each decoding step. While speculative decoding is a widely accepted technique to accelerate autoregressive decoding, existing methods often struggle to achieve significant speedups due to inefficient KV cache optimization strategies and result in low acceptance rates. To a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.10424","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/2502.10424/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:14:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JG7DT2DhlICxwyUvvdHFu6XEu5X+/6SjIBOHdNm6YwjcmWVnWxbIGdcVsUZuk5v2yKj2yvfk2zfknRnBLgaPBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T01:21:36.712699Z"},"content_sha256":"34479c1e2a770edd97c0de3f9348fd1543b19ce70cfecdbf17573ed6db9ca652","schema_version":"1.0","event_id":"sha256:34479c1e2a770edd97c0de3f9348fd1543b19ce70cfecdbf17573ed6db9ca652"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MJJN7I5Q5KHIAK2DJHDEK4OL2Y/bundle.json","state_url":"https://pith.science/pith/MJJN7I5Q5KHIAK2DJHDEK4OL2Y/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MJJN7I5Q5KHIAK2DJHDEK4OL2Y/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-19T01:21:36Z","links":{"resolver":"https://pith.science/pith/MJJN7I5Q5KHIAK2DJHDEK4OL2Y","bundle":"https://pith.science/pith/MJJN7I5Q5KHIAK2DJHDEK4OL2Y/bundle.json","state":"https://pith.science/pith/MJJN7I5Q5KHIAK2DJHDEK4OL2Y/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MJJN7I5Q5KHIAK2DJHDEK4OL2Y/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:MJJN7I5Q5KHIAK2DJHDEK4OL2Y","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"59a161c652e7cd125bb52f1a53dd9e2511918b30861cd11d953e38873cbfc7cf","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-05T20:43:48Z","title_canon_sha256":"1f2e959a2931f5d5a25dc129795dbb0ea2f24fef1056563c44d12f48b33ea13e"},"schema_version":"1.0","source":{"id":"2502.10424","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.10424","created_at":"2026-07-05T10:14:52Z"},{"alias_kind":"arxiv_version","alias_value":"2502.10424v1","created_at":"2026-07-05T10:14:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.10424","created_at":"2026-07-05T10:14:52Z"},{"alias_kind":"pith_short_12","alias_value":"MJJN7I5Q5KHI","created_at":"2026-07-05T10:14:52Z"},{"alias_kind":"pith_short_16","alias_value":"MJJN7I5Q5KHIAK2D","created_at":"2026-07-05T10:14:52Z"},{"alias_kind":"pith_short_8","alias_value":"MJJN7I5Q","created_at":"2026-07-05T10:14:52Z"}],"graph_snapshots":[{"event_id":"sha256:34479c1e2a770edd97c0de3f9348fd1543b19ce70cfecdbf17573ed6db9ca652","target":"graph","created_at":"2026-07-05T10:14:52Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2502.10424/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) are increasingly being deployed on edge devices for long-context settings, creating a growing need for fast and efficient long-context inference. In these scenarios, the Key-Value (KV) cache is the primary bottleneck in terms of both GPU memory and latency, as the full KV cache must be loaded for each decoding step. While speculative decoding is a widely accepted technique to accelerate autoregressive decoding, existing methods often struggle to achieve significant speedups due to inefficient KV cache optimization strategies and result in low acceptance rates. To a","authors_text":"Aditya Tomar, Amir Gholami, Coleman Hooper, Haocheng Xi, Kurt Keutzer, Mahyar Najibi, Maxwell Horton, Michael W. Mahoney, Rishabh Tiwari, Sehoon Kim","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-05T20:43:48Z","title":"QuantSpec: Self-Speculative Decoding with Hierarchical Quantized KV Cache"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.10424","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:aed677611980c6b5b152a76e018a1214ca60dfcba45f3e5d08538760ed981410","target":"record","created_at":"2026-07-05T10:14:52Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"59a161c652e7cd125bb52f1a53dd9e2511918b30861cd11d953e38873cbfc7cf","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-05T20:43:48Z","title_canon_sha256":"1f2e959a2931f5d5a25dc129795dbb0ea2f24fef1056563c44d12f48b33ea13e"},"schema_version":"1.0","source":{"id":"2502.10424","kind":"arxiv","version":1}},"canonical_sha256":"6252dfa3b0ea8e802b4349c64571cbd6320333add6751f802b76f17074b8514a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6252dfa3b0ea8e802b4349c64571cbd6320333add6751f802b76f17074b8514a","first_computed_at":"2026-07-05T10:14:52.380637Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:14:52.380637Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"sWEx/zeyUmeEZ1Vb2F+dJNRpokZukxB9GZxnEiKNd2wrQFymMnYZbTaKFKde1eSSB5m2zMvgUZy0/ENNet+1DQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:14:52.381129Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.10424","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:aed677611980c6b5b152a76e018a1214ca60dfcba45f3e5d08538760ed981410","sha256:34479c1e2a770edd97c0de3f9348fd1543b19ce70cfecdbf17573ed6db9ca652"],"state_sha256":"8971ebce93ecf78019cf39a24b262b0836e1049b2fc70e3f129adf5c79b392bf"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"doth3CJxabz3aRV4MCkwIO+BZ3lJoWrScH39IFSIvEIHTUHLHhrV5guLTMMSSsQkEFN9WKXwaDAJDJR7yLIzCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-19T01:21:36.718489Z","bundle_sha256":"9d68576e0d5fa18a61e03b32bf67642ad6b8ed1f34ea1dfe9e7b9c96fdf82240"}}