{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:AX3F35SMSJVET5CFQM34OXYYJM","short_pith_number":"pith:AX3F35SM","schema_version":"1.0","canonical_sha256":"05f65df64c926a49f4458337c75f184b1a73f6ba38429112dc59be95f5289248","source":{"kind":"arxiv","id":"2506.14158","version":1},"attestation_state":"computed","paper":{"title":"S$^4$C: Speculative Sampling with Syntactic and Semantic Coherence for Efficient Inference of Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Feng Tian, Guang Huang, Guiguang Ding, Pengyang Wang, Sicheng Zhao, Tao He, Tianshi Xu, Yu Yang","submitted_at":"2025-06-17T03:38:19Z","abstract_excerpt":"Large language models (LLMs) exhibit remarkable reasoning capabilities across diverse downstream tasks. However, their autoregressive nature leads to substantial inference latency, posing challenges for real-time applications. Speculative sampling mitigates this issue by introducing a drafting phase followed by a parallel validation phase, enabling faster token generation and verification. Existing approaches, however, overlook the inherent coherence in text generation, limiting their efficiency. To address this gap, we propose a Speculative Sampling with Syntactic and Semantic Coherence (S$^4"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2506.14158","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-17T03:38:19Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a11b3fad26978343be4fb7412c903e9ee180d5032a14608f9ca9308ca4617d54","abstract_canon_sha256":"ce8fe0eb5dc355e488dae91518949cb909f90e02e25cefb5eb11591b3dcefa27"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:22:50.104733Z","signature_b64":"BwEf5q+inOzVUhr3xbKbsghajbGbrje2ehIBZzS1KewA+KjB2vEdAIcFQtJbgJK9Ps3zPRGJMnrUoU99eN7SBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"05f65df64c926a49f4458337c75f184b1a73f6ba38429112dc59be95f5289248","last_reissued_at":"2026-07-05T11:22:50.104251Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:22:50.104251Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"S$^4$C: Speculative Sampling with Syntactic and Semantic Coherence for Efficient Inference of Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Feng Tian, Guang Huang, Guiguang Ding, Pengyang Wang, Sicheng Zhao, Tao He, Tianshi Xu, Yu Yang","submitted_at":"2025-06-17T03:38:19Z","abstract_excerpt":"Large language models (LLMs) exhibit remarkable reasoning capabilities across diverse downstream tasks. However, their autoregressive nature leads to substantial inference latency, posing challenges for real-time applications. Speculative sampling mitigates this issue by introducing a drafting phase followed by a parallel validation phase, enabling faster token generation and verification. Existing approaches, however, overlook the inherent coherence in text generation, limiting their efficiency. To address this gap, we propose a Speculative Sampling with Syntactic and Semantic Coherence (S$^4"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.14158","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/2506.14158/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":"2506.14158","created_at":"2026-07-05T11:22:50.104302+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.14158v1","created_at":"2026-07-05T11:22:50.104302+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.14158","created_at":"2026-07-05T11:22:50.104302+00:00"},{"alias_kind":"pith_short_12","alias_value":"AX3F35SMSJVE","created_at":"2026-07-05T11:22:50.104302+00:00"},{"alias_kind":"pith_short_16","alias_value":"AX3F35SMSJVET5CF","created_at":"2026-07-05T11:22:50.104302+00:00"},{"alias_kind":"pith_short_8","alias_value":"AX3F35SM","created_at":"2026-07-05T11:22:50.104302+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/AX3F35SMSJVET5CFQM34OXYYJM","json":"https://pith.science/pith/AX3F35SMSJVET5CFQM34OXYYJM.json","graph_json":"https://pith.science/api/pith-number/AX3F35SMSJVET5CFQM34OXYYJM/graph.json","events_json":"https://pith.science/api/pith-number/AX3F35SMSJVET5CFQM34OXYYJM/events.json","paper":"https://pith.science/paper/AX3F35SM"},"agent_actions":{"view_html":"https://pith.science/pith/AX3F35SMSJVET5CFQM34OXYYJM","download_json":"https://pith.science/pith/AX3F35SMSJVET5CFQM34OXYYJM.json","view_paper":"https://pith.science/paper/AX3F35SM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.14158&json=true","fetch_graph":"https://pith.science/api/pith-number/AX3F35SMSJVET5CFQM34OXYYJM/graph.json","fetch_events":"https://pith.science/api/pith-number/AX3F35SMSJVET5CFQM34OXYYJM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AX3F35SMSJVET5CFQM34OXYYJM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AX3F35SMSJVET5CFQM34OXYYJM/action/storage_attestation","attest_author":"https://pith.science/pith/AX3F35SMSJVET5CFQM34OXYYJM/action/author_attestation","sign_citation":"https://pith.science/pith/AX3F35SMSJVET5CFQM34OXYYJM/action/citation_signature","submit_replication":"https://pith.science/pith/AX3F35SMSJVET5CFQM34OXYYJM/action/replication_record"}},"created_at":"2026-07-05T11:22:50.104302+00:00","updated_at":"2026-07-05T11:22:50.104302+00:00"}