{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:A35AM6EABHLQKT3ZTYANKQJPQN","short_pith_number":"pith:A35AM6EA","canonical_record":{"source":{"id":"2602.06036","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-05T18:59:30Z","cross_cats_sorted":[],"title_canon_sha256":"0e8334084a7f240bb4426e8ac325e61b82231790084ea4bd100ac3de18d7e6c4","abstract_canon_sha256":"09bd5f9fb406c5da53efae0dfe7353b7eb95a304a727de7ed6e4703fa5705e71"},"schema_version":"1.0"},"canonical_sha256":"06fa06788009d7054f799e00d5412f836ca0927544fa44c74863baf5d51d68d3","source":{"kind":"arxiv","id":"2602.06036","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.06036","created_at":"2026-05-29T01:05:03Z"},{"alias_kind":"arxiv_version","alias_value":"2602.06036v2","created_at":"2026-05-29T01:05:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.06036","created_at":"2026-05-29T01:05:03Z"},{"alias_kind":"pith_short_12","alias_value":"A35AM6EABHLQ","created_at":"2026-05-29T01:05:03Z"},{"alias_kind":"pith_short_16","alias_value":"A35AM6EABHLQKT3Z","created_at":"2026-05-29T01:05:03Z"},{"alias_kind":"pith_short_8","alias_value":"A35AM6EA","created_at":"2026-05-29T01:05:03Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:A35AM6EABHLQKT3ZTYANKQJPQN","target":"record","payload":{"canonical_record":{"source":{"id":"2602.06036","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-05T18:59:30Z","cross_cats_sorted":[],"title_canon_sha256":"0e8334084a7f240bb4426e8ac325e61b82231790084ea4bd100ac3de18d7e6c4","abstract_canon_sha256":"09bd5f9fb406c5da53efae0dfe7353b7eb95a304a727de7ed6e4703fa5705e71"},"schema_version":"1.0"},"canonical_sha256":"06fa06788009d7054f799e00d5412f836ca0927544fa44c74863baf5d51d68d3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T01:05:03.617081Z","signature_b64":"Vc3VCBaicolz7sOcnq1b8tGBk4MdWABIHB/zpiX/xu7hq4FT5pO5jtK6RLO/r2NEdGxkoPWyx/dsvn8pDVT0DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"06fa06788009d7054f799e00d5412f836ca0927544fa44c74863baf5d51d68d3","last_reissued_at":"2026-05-29T01:05:03.616090Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T01:05:03.616090Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2602.06036","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-29T01:05:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zQ29NLxVr+inEnDn3CUrzcmli/tpvXrLhg286iK49+S5dHEGJRtVx4rxQDYEBqgKdFZ3RFZoqDr9KR1tf/rFDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-29T22:30:26.057581Z"},"content_sha256":"697308ab9af7b60cccef35a34a4567380fc4e208f2ba5c64fe0e8320225f502d","schema_version":"1.0","event_id":"sha256:697308ab9af7b60cccef35a34a4567380fc4e208f2ba5c64fe0e8320225f502d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:A35AM6EABHLQKT3ZTYANKQJPQN","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"DFlash: Block Diffusion for Flash Speculative Decoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jian Chen, Yesheng Liang, Zhijian Liu","submitted_at":"2026-02-05T18:59:30Z","abstract_excerpt":"Autoregressive large language models (LLMs) deliver strong performance but require inherently sequential decoding, leading to high inference latency and poor GPU utilization. Speculative decoding mitigates this bottleneck by using a fast draft model whose outputs are verified in parallel by the target LLM; however, existing methods still rely on autoregressive drafting, which remains sequential and limits practical speedups. Diffusion LLMs offer a promising alternative by enabling parallel generation, but current diffusion models typically underperform compared with autoregressive models. In t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.06036","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/2602.06036/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":"2026-05-29T01:05:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"P1UXd4vK0cbfJiAhWiXQ4ApwKAIcMvvqVJFyDrsgWxcsYuzddnDfMoxR2Hkn2l8LjixQXwN/wDz33L4MHsDoDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-29T22:30:26.058245Z"},"content_sha256":"c7692b98b37616f074abccd58ef1c1875045187494115ace1bce0a24a6c6a2c6","schema_version":"1.0","event_id":"sha256:c7692b98b37616f074abccd58ef1c1875045187494115ace1bce0a24a6c6a2c6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/A35AM6EABHLQKT3ZTYANKQJPQN/bundle.json","state_url":"https://pith.science/pith/A35AM6EABHLQKT3ZTYANKQJPQN/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/A35AM6EABHLQKT3ZTYANKQJPQN/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":"2026-05-29T22:30:26Z","links":{"resolver":"https://pith.science/pith/A35AM6EABHLQKT3ZTYANKQJPQN","bundle":"https://pith.science/pith/A35AM6EABHLQKT3ZTYANKQJPQN/bundle.json","state":"https://pith.science/pith/A35AM6EABHLQKT3ZTYANKQJPQN/state.json","well_known_bundle":"https://pith.science/.well-known/pith/A35AM6EABHLQKT3ZTYANKQJPQN/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:A35AM6EABHLQKT3ZTYANKQJPQN","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":"09bd5f9fb406c5da53efae0dfe7353b7eb95a304a727de7ed6e4703fa5705e71","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-05T18:59:30Z","title_canon_sha256":"0e8334084a7f240bb4426e8ac325e61b82231790084ea4bd100ac3de18d7e6c4"},"schema_version":"1.0","source":{"id":"2602.06036","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.06036","created_at":"2026-05-29T01:05:03Z"},{"alias_kind":"arxiv_version","alias_value":"2602.06036v2","created_at":"2026-05-29T01:05:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.06036","created_at":"2026-05-29T01:05:03Z"},{"alias_kind":"pith_short_12","alias_value":"A35AM6EABHLQ","created_at":"2026-05-29T01:05:03Z"},{"alias_kind":"pith_short_16","alias_value":"A35AM6EABHLQKT3Z","created_at":"2026-05-29T01:05:03Z"},{"alias_kind":"pith_short_8","alias_value":"A35AM6EA","created_at":"2026-05-29T01:05:03Z"}],"graph_snapshots":[{"event_id":"sha256:c7692b98b37616f074abccd58ef1c1875045187494115ace1bce0a24a6c6a2c6","target":"graph","created_at":"2026-05-29T01:05:03Z","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/2602.06036/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Autoregressive large language models (LLMs) deliver strong performance but require inherently sequential decoding, leading to high inference latency and poor GPU utilization. Speculative decoding mitigates this bottleneck by using a fast draft model whose outputs are verified in parallel by the target LLM; however, existing methods still rely on autoregressive drafting, which remains sequential and limits practical speedups. Diffusion LLMs offer a promising alternative by enabling parallel generation, but current diffusion models typically underperform compared with autoregressive models. In t","authors_text":"Jian Chen, Yesheng Liang, Zhijian Liu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-05T18:59:30Z","title":"DFlash: Block Diffusion for Flash Speculative Decoding"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.06036","kind":"arxiv","version":2},"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":"sha256:697308ab9af7b60cccef35a34a4567380fc4e208f2ba5c64fe0e8320225f502d","target":"record","created_at":"2026-05-29T01:05:03Z","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":"09bd5f9fb406c5da53efae0dfe7353b7eb95a304a727de7ed6e4703fa5705e71","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-05T18:59:30Z","title_canon_sha256":"0e8334084a7f240bb4426e8ac325e61b82231790084ea4bd100ac3de18d7e6c4"},"schema_version":"1.0","source":{"id":"2602.06036","kind":"arxiv","version":2}},"canonical_sha256":"06fa06788009d7054f799e00d5412f836ca0927544fa44c74863baf5d51d68d3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"06fa06788009d7054f799e00d5412f836ca0927544fa44c74863baf5d51d68d3","first_computed_at":"2026-05-29T01:05:03.616090Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-29T01:05:03.616090Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Vc3VCBaicolz7sOcnq1b8tGBk4MdWABIHB/zpiX/xu7hq4FT5pO5jtK6RLO/r2NEdGxkoPWyx/dsvn8pDVT0DQ==","signature_status":"signed_v1","signed_at":"2026-05-29T01:05:03.617081Z","signed_message":"canonical_sha256_bytes"},"source_id":"2602.06036","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:697308ab9af7b60cccef35a34a4567380fc4e208f2ba5c64fe0e8320225f502d","sha256:c7692b98b37616f074abccd58ef1c1875045187494115ace1bce0a24a6c6a2c6"],"state_sha256":"70e900756caf2d727bd87d13cbea0ae2dc041874419b1e655e6cc8a53dc2387c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"DIwkKPYMfQdtDDawk/6Sxp+gTANlfdAfrPs3SvjrgMLLLmU6CrW5EYA+sMOTUhmRWyS5dkvQTMJJc2MSVlDSCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-29T22:30:26.062247Z","bundle_sha256":"bc6f453f35b341694d2cd4d5eefd5619ac486cc26e39888c3d4b6f64a06f0dcc"}}