{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:QKZG227XCNL6JLG6YHNGFZCSHU","short_pith_number":"pith:QKZG227X","schema_version":"1.0","canonical_sha256":"82b26d6bf71357e4acdec1da62e4523d195880b192c8adce9c600148aa736334","source":{"kind":"arxiv","id":"2507.05687","version":1},"attestation_state":"computed","paper":{"title":"AutoTriton: Automatic Triton Programming with Reinforcement Learning in LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Jianling Li, Maosong Sun, Qi Shi, Shangzhan Li, Wanxiang Che, Xu Han, Ye He, Yonggang Hu, Yuxuan Li, Zefan Wang, Zhiyuan Liu","submitted_at":"2025-07-08T05:38:24Z","abstract_excerpt":"Kernel development in deep learning requires optimizing computational units across hardware while balancing memory management, parallelism, and hardware-specific optimizations through extensive empirical tuning. Although domain-specific languages like Triton simplify GPU programming by abstracting low-level details, developers must still manually tune critical parameters such as tile sizes and memory access patterns through iterative experimentation, creating substantial barriers to optimal performance and wider adoption. In this work, we introduce AutoTriton, the first model dedicated to Trit"},"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":"2507.05687","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-07-08T05:38:24Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"9f1d97ee7f1fabb244876051ddf697996ae7cffef05ba615fee5d74f49a77e37","abstract_canon_sha256":"1952b8b791f2ab4f4aaa68759d9e50dd0d30e8ec6d64bd9db61a464d76ec00f9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:33:34.859800Z","signature_b64":"yucc6H4zikFB4L+4jT5KmjzuuaCxG2AxvjS8eFMtGNcN4+R7AnhSJ9Y9VIyEowGBjZ9l4h2Hks2FBPh8b356Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"82b26d6bf71357e4acdec1da62e4523d195880b192c8adce9c600148aa736334","last_reissued_at":"2026-07-05T11:33:34.859338Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:33:34.859338Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AutoTriton: Automatic Triton Programming with Reinforcement Learning in LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Jianling Li, Maosong Sun, Qi Shi, Shangzhan Li, Wanxiang Che, Xu Han, Ye He, Yonggang Hu, Yuxuan Li, Zefan Wang, Zhiyuan Liu","submitted_at":"2025-07-08T05:38:24Z","abstract_excerpt":"Kernel development in deep learning requires optimizing computational units across hardware while balancing memory management, parallelism, and hardware-specific optimizations through extensive empirical tuning. Although domain-specific languages like Triton simplify GPU programming by abstracting low-level details, developers must still manually tune critical parameters such as tile sizes and memory access patterns through iterative experimentation, creating substantial barriers to optimal performance and wider adoption. In this work, we introduce AutoTriton, the first model dedicated to Trit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.05687","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/2507.05687/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":"2507.05687","created_at":"2026-07-05T11:33:34.859396+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.05687v1","created_at":"2026-07-05T11:33:34.859396+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.05687","created_at":"2026-07-05T11:33:34.859396+00:00"},{"alias_kind":"pith_short_12","alias_value":"QKZG227XCNL6","created_at":"2026-07-05T11:33:34.859396+00:00"},{"alias_kind":"pith_short_16","alias_value":"QKZG227XCNL6JLG6","created_at":"2026-07-05T11:33:34.859396+00:00"},{"alias_kind":"pith_short_8","alias_value":"QKZG227X","created_at":"2026-07-05T11:33:34.859396+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":11,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.16497","citing_title":"daVinci-kernel: Co-Evolving Skill Selection, Summarization, and Utilization via RL for GPU Kernel Optimization","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30359","citing_title":"Kernel Foundry: A Diagnosis-driven Evolutionary Kernel Optimizer with Multi-Experts","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28213","citing_title":"Learning When to Optimize: Verified Optimization Skills from Expert GPU-Kernel Lineages","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2606.04847","citing_title":"MusaCoder: Native GPU Kernel Generation with Full-Stack Training on Moore Threads GPU","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23215","citing_title":"FastKernels: Benchmarking GPU Kernel Generation in Production","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2603.23566","citing_title":"AscendOptimizer: Episodic Agent for Ascend NPU Operator Optimization","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17978","citing_title":"AutoVecCoder: Teaching LLMs to Generate Explicitly Vectorized Code","ref_index":76,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16819","citing_title":"AgentKernelArena: Generalization-Aware Benchmarking of GPU Kernel Optimization Agents","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2603.28342","citing_title":"Kernel-Smith: A Unified Recipe for Evolutionary Kernel Optimization","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04956","citing_title":"KernelBenchX: A Comprehensive Benchmark for Evaluating LLM-Generated GPU Kernels","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04956","citing_title":"KernelBenchX: A Comprehensive Benchmark for Evaluating LLM-Generated GPU Kernels","ref_index":5,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QKZG227XCNL6JLG6YHNGFZCSHU","json":"https://pith.science/pith/QKZG227XCNL6JLG6YHNGFZCSHU.json","graph_json":"https://pith.science/api/pith-number/QKZG227XCNL6JLG6YHNGFZCSHU/graph.json","events_json":"https://pith.science/api/pith-number/QKZG227XCNL6JLG6YHNGFZCSHU/events.json","paper":"https://pith.science/paper/QKZG227X"},"agent_actions":{"view_html":"https://pith.science/pith/QKZG227XCNL6JLG6YHNGFZCSHU","download_json":"https://pith.science/pith/QKZG227XCNL6JLG6YHNGFZCSHU.json","view_paper":"https://pith.science/paper/QKZG227X","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.05687&json=true","fetch_graph":"https://pith.science/api/pith-number/QKZG227XCNL6JLG6YHNGFZCSHU/graph.json","fetch_events":"https://pith.science/api/pith-number/QKZG227XCNL6JLG6YHNGFZCSHU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QKZG227XCNL6JLG6YHNGFZCSHU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QKZG227XCNL6JLG6YHNGFZCSHU/action/storage_attestation","attest_author":"https://pith.science/pith/QKZG227XCNL6JLG6YHNGFZCSHU/action/author_attestation","sign_citation":"https://pith.science/pith/QKZG227XCNL6JLG6YHNGFZCSHU/action/citation_signature","submit_replication":"https://pith.science/pith/QKZG227XCNL6JLG6YHNGFZCSHU/action/replication_record"}},"created_at":"2026-07-05T11:33:34.859396+00:00","updated_at":"2026-07-05T11:33:34.859396+00:00"}