{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:MAP6H32XF56M4LKG35IOKMVO5Y","short_pith_number":"pith:MAP6H32X","schema_version":"1.0","canonical_sha256":"601fe3ef572f7cce2d46df50e532aeee1e38f10fdf7d45e4269d1471f3b83493","source":{"kind":"arxiv","id":"2605.30851","version":1},"attestation_state":"computed","paper":{"title":"How Much Parallelism Is \"Free\"? A Principle of Near-Free Parallelism for Parallel Decoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.PF","authors_text":"Aiwei Liu, Lingzhe Zhang, Minghua He, Xiao Zhou, Yuan Liu","submitted_at":"2026-05-29T05:16:12Z","abstract_excerpt":"Parallel decoding improves generation efficiency by processing multiple decode positions within a single decode forward, but reported speedups conflate algorithmic token utilization with the system cost of executing multiple positions. We isolate the system side by introducing Near-Free Parallelism (NFP), the maximum number of positions executable at near-free latency. Analyzing Dense FFNs, MoE FFNs, and Attention against an idle-compute baseline, we find that NFP is shaped not by memory-bound resource slack alone, but also by implementation-induced kernel-granularity slack. Based on these mec"},"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":"2605.30851","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.PF","submitted_at":"2026-05-29T05:16:12Z","cross_cats_sorted":[],"title_canon_sha256":"51931d0e7a602748df24f6e315a8402117cbfb08800f18491c87c01085458685","abstract_canon_sha256":"de197192510260e15446f4bde60765797e954221c18cb0e641c9cad9bd64b491"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T01:03:21.016445Z","signature_b64":"+xQ14hxdN9axJvZ5ImXlf9rLfFVfyqkCgqS7Mr109m7DYR16ndIjyOY3+PgqeZXAtow1w81lx41ljaUtOeP0CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"601fe3ef572f7cce2d46df50e532aeee1e38f10fdf7d45e4269d1471f3b83493","last_reissued_at":"2026-06-01T01:03:21.015578Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T01:03:21.015578Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How Much Parallelism Is \"Free\"? A Principle of Near-Free Parallelism for Parallel Decoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.PF","authors_text":"Aiwei Liu, Lingzhe Zhang, Minghua He, Xiao Zhou, Yuan Liu","submitted_at":"2026-05-29T05:16:12Z","abstract_excerpt":"Parallel decoding improves generation efficiency by processing multiple decode positions within a single decode forward, but reported speedups conflate algorithmic token utilization with the system cost of executing multiple positions. We isolate the system side by introducing Near-Free Parallelism (NFP), the maximum number of positions executable at near-free latency. Analyzing Dense FFNs, MoE FFNs, and Attention against an idle-compute baseline, we find that NFP is shaped not by memory-bound resource slack alone, but also by implementation-induced kernel-granularity slack. Based on these mec"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30851","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/2605.30851/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":"2605.30851","created_at":"2026-06-01T01:03:21.015690+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.30851v1","created_at":"2026-06-01T01:03:21.015690+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30851","created_at":"2026-06-01T01:03:21.015690+00:00"},{"alias_kind":"pith_short_12","alias_value":"MAP6H32XF56M","created_at":"2026-06-01T01:03:21.015690+00:00"},{"alias_kind":"pith_short_16","alias_value":"MAP6H32XF56M4LKG","created_at":"2026-06-01T01:03:21.015690+00:00"},{"alias_kind":"pith_short_8","alias_value":"MAP6H32X","created_at":"2026-06-01T01:03:21.015690+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/MAP6H32XF56M4LKG35IOKMVO5Y","json":"https://pith.science/pith/MAP6H32XF56M4LKG35IOKMVO5Y.json","graph_json":"https://pith.science/api/pith-number/MAP6H32XF56M4LKG35IOKMVO5Y/graph.json","events_json":"https://pith.science/api/pith-number/MAP6H32XF56M4LKG35IOKMVO5Y/events.json","paper":"https://pith.science/paper/MAP6H32X"},"agent_actions":{"view_html":"https://pith.science/pith/MAP6H32XF56M4LKG35IOKMVO5Y","download_json":"https://pith.science/pith/MAP6H32XF56M4LKG35IOKMVO5Y.json","view_paper":"https://pith.science/paper/MAP6H32X","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.30851&json=true","fetch_graph":"https://pith.science/api/pith-number/MAP6H32XF56M4LKG35IOKMVO5Y/graph.json","fetch_events":"https://pith.science/api/pith-number/MAP6H32XF56M4LKG35IOKMVO5Y/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MAP6H32XF56M4LKG35IOKMVO5Y/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MAP6H32XF56M4LKG35IOKMVO5Y/action/storage_attestation","attest_author":"https://pith.science/pith/MAP6H32XF56M4LKG35IOKMVO5Y/action/author_attestation","sign_citation":"https://pith.science/pith/MAP6H32XF56M4LKG35IOKMVO5Y/action/citation_signature","submit_replication":"https://pith.science/pith/MAP6H32XF56M4LKG35IOKMVO5Y/action/replication_record"}},"created_at":"2026-06-01T01:03:21.015690+00:00","updated_at":"2026-06-01T01:03:21.015690+00:00"}