{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:REYUUESQYWKUETQFO32U2CAYDB","short_pith_number":"pith:REYUUESQ","schema_version":"1.0","canonical_sha256":"89314a1250c595424e0576f54d0818184e254bf3489a7dc0aff2bf814041382a","source":{"kind":"arxiv","id":"2606.17518","version":1},"attestation_state":"computed","paper":{"title":"SpecGen: Accelerating Agentic Kernel Optimization with Speculative Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Dahua Lin, Jihu Guo, Sitian Lu, Tenghui Ma, Wei Gao, Xingcheng Zhang, Zhisheng Ye","submitted_at":"2026-06-16T04:54:23Z","abstract_excerpt":"Agentic kernel optimization automates manual GPU kernel tuning via iterative generation, validation, and profiling with reasoning LLMs, casting the optimization task as feedback-guided search. However, our workload characterization reveals three system-level inefficiencies that limit search efficiency: (1) long generation latency due to LLM reasoning, (2) insufficient profiling feedback, and (3) underutilized validation/profiling resources. Our key insight is that the ongoing reasoning generation exposes a window for producing additional candidate kernels before it completes, allowing the syst"},"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":"2606.17518","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-06-16T04:54:23Z","cross_cats_sorted":[],"title_canon_sha256":"a2428e0722376924e71632363af527e11221cee7b289605145a6adc645ab0558","abstract_canon_sha256":"b56261f11770fcdd042daca590a691db67c3b7904031cf1e6681b07ed24569f3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:10:14.963498Z","signature_b64":"BQRR7OaPwduBPly9T9b2vxvi6LxcI+4IYG5DUkcNXZhIvRjdpNhb5frjct9N30hcms+zWdN6HzKWo5x1jqeSAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"89314a1250c595424e0576f54d0818184e254bf3489a7dc0aff2bf814041382a","last_reissued_at":"2026-06-19T16:10:14.963123Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:10:14.963123Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SpecGen: Accelerating Agentic Kernel Optimization with Speculative Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Dahua Lin, Jihu Guo, Sitian Lu, Tenghui Ma, Wei Gao, Xingcheng Zhang, Zhisheng Ye","submitted_at":"2026-06-16T04:54:23Z","abstract_excerpt":"Agentic kernel optimization automates manual GPU kernel tuning via iterative generation, validation, and profiling with reasoning LLMs, casting the optimization task as feedback-guided search. However, our workload characterization reveals three system-level inefficiencies that limit search efficiency: (1) long generation latency due to LLM reasoning, (2) insufficient profiling feedback, and (3) underutilized validation/profiling resources. Our key insight is that the ongoing reasoning generation exposes a window for producing additional candidate kernels before it completes, allowing the syst"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.17518","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/2606.17518/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":"2606.17518","created_at":"2026-06-19T16:10:14.963187+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.17518v1","created_at":"2026-06-19T16:10:14.963187+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.17518","created_at":"2026-06-19T16:10:14.963187+00:00"},{"alias_kind":"pith_short_12","alias_value":"REYUUESQYWKU","created_at":"2026-06-19T16:10:14.963187+00:00"},{"alias_kind":"pith_short_16","alias_value":"REYUUESQYWKUETQF","created_at":"2026-06-19T16:10:14.963187+00:00"},{"alias_kind":"pith_short_8","alias_value":"REYUUESQ","created_at":"2026-06-19T16:10:14.963187+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/REYUUESQYWKUETQFO32U2CAYDB","json":"https://pith.science/pith/REYUUESQYWKUETQFO32U2CAYDB.json","graph_json":"https://pith.science/api/pith-number/REYUUESQYWKUETQFO32U2CAYDB/graph.json","events_json":"https://pith.science/api/pith-number/REYUUESQYWKUETQFO32U2CAYDB/events.json","paper":"https://pith.science/paper/REYUUESQ"},"agent_actions":{"view_html":"https://pith.science/pith/REYUUESQYWKUETQFO32U2CAYDB","download_json":"https://pith.science/pith/REYUUESQYWKUETQFO32U2CAYDB.json","view_paper":"https://pith.science/paper/REYUUESQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.17518&json=true","fetch_graph":"https://pith.science/api/pith-number/REYUUESQYWKUETQFO32U2CAYDB/graph.json","fetch_events":"https://pith.science/api/pith-number/REYUUESQYWKUETQFO32U2CAYDB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/REYUUESQYWKUETQFO32U2CAYDB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/REYUUESQYWKUETQFO32U2CAYDB/action/storage_attestation","attest_author":"https://pith.science/pith/REYUUESQYWKUETQFO32U2CAYDB/action/author_attestation","sign_citation":"https://pith.science/pith/REYUUESQYWKUETQFO32U2CAYDB/action/citation_signature","submit_replication":"https://pith.science/pith/REYUUESQYWKUETQFO32U2CAYDB/action/replication_record"}},"created_at":"2026-06-19T16:10:14.963187+00:00","updated_at":"2026-06-19T16:10:14.963187+00:00"}