{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/arxiv-2604-16625","title":"AdaExplore: Failure-Driven Adaptation and Diversity-Preserving Search for Efficient Kernel Generation","arxiv_id":"2604.16625","date":"2026-04-17","proceeding":null,"authors":["Weihua Du","Jingming Zhuo","Yixin Dong","Andre Wang He","Weiwei Sun","Zeyu Zheng","Manupa Karunaratne","Ivan Fox","Tim Dettmers","Tianqi Chen","Yiming Yang","Sean Welleck"],"abstract":"Recent large language model (LLM) agents have shown promise in using execution feedback for test-time adaptation. However, robust self-improvement remains far from solved: most approaches still treat each problem instance independently, without accumulating reusable knowledge. This limitation is particularly pronounced in domain-specific languages such as Triton, which are underrepresented in LLM pretraining data. Their strict constraints and non-linear optimization landscape further make naive generation and local refinement unreliable. We propose AdaExplore, an agent framework that enables self-improvement via accumulated execution feedback for performance-critical kernel code generation through two complementary stages: failure-driven adaptation and diversity-preserving search, jointly improving correctness and optimization performance without additional fine-tuning or external knowledge. In the adaptation stage, the agent synthesizes tasks and converts recurring failures into a reusable memory of validity rules, helping subsequent generations remain within the feasible set. In the search stage, the agent organizes candidate kernels as a tree and alternates between small local refinements and larger structural regeneration, allowing it to explore the optimization landscape beyond local optima. Experiments on kernel runtime optimization benchmarks validate these gains: AdaExplore achieves 3.12x and 1.72x speedups on KernelBench Level-2 and Level-3, respectively, within 100 steps, and continues to improve with additional computation.","url_abs":"https://arxiv.org/abs/2604.16625","url_pdf":"https://arxiv.org/pdf/2604.16625","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2604.16625","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2604.16625"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/StigLidu/AdaExplore","reach":null}],"summary":{"unverified":9},"by_repo_kind":{"found_in_text":{"samples":9,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"57158d2c929b3690","entry":"construct_kernelbench_dataset","repo":"StigLidu/AdaExplore","repo_kind":"found_in_text","path":"src/dataset.py","file_url":"https://github.com/StigLidu/AdaExplore/blob/HEAD/src/dataset.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"57158d2c929b3690"}},{"code_sha256_prefix":"e3b75273924ba2be","entry":"construct_problem_dataset_from_problem_dir","repo":"StigLidu/AdaExplore","repo_kind":"found_in_text","path":"src/dataset.py","file_url":"https://github.com/StigLidu/AdaExplore/blob/HEAD/src/dataset.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e3b75273924ba2be"}},{"code_sha256_prefix":"9c9ffe93baf02d12","entry":"dummy_small_step","repo":"StigLidu/AdaExplore","repo_kind":"found_in_text","path":"agent/actions.py","file_url":"https://github.com/StigLidu/AdaExplore/blob/HEAD/agent/actions.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9c9ffe93baf02d12"}},{"code_sha256_prefix":"b85d265b7395960b","entry":"extract_first_code","repo":"StigLidu/AdaExplore","repo_kind":"found_in_text","path":"src/utils.py","file_url":"https://github.com/StigLidu/AdaExplore/blob/HEAD/src/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b85d265b7395960b"}},{"code_sha256_prefix":"2e5d7d02ef0d1a63","entry":"generate_experience_guidance_prompt","repo":"StigLidu/AdaExplore","repo_kind":"found_in_text","path":"agentprompt/prompt_modules.py","file_url":"https://github.com/StigLidu/AdaExplore/blob/HEAD/agentprompt/prompt_modules.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2e5d7d02ef0d1a63"}},{"code_sha256_prefix":"0b8e26831e229f55","entry":"generate_hardware_information_prompt","repo":"StigLidu/AdaExplore","repo_kind":"found_in_text","path":"agentprompt/prompt_modules.py","file_url":"https://github.com/StigLidu/AdaExplore/blob/HEAD/agentprompt/prompt_modules.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0b8e26831e229f55"}},{"code_sha256_prefix":"0346a0fb1539052c","entry":"get_code_hash","repo":"StigLidu/AdaExplore","repo_kind":"found_in_text","path":"src/dataset.py","file_url":"https://github.com/StigLidu/AdaExplore/blob/HEAD/src/dataset.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0346a0fb1539052c"}},{"code_sha256_prefix":"aad3c512cb2a7994","entry":"query_inference_server","repo":"StigLidu/AdaExplore","repo_kind":"found_in_text","path":"agent/inference_server.py","file_url":"https://github.com/StigLidu/AdaExplore/blob/HEAD/agent/inference_server.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"aad3c512cb2a7994"}},{"code_sha256_prefix":"20ac9f8ab3912be1","entry":"read_file","repo":"StigLidu/AdaExplore","repo_kind":"found_in_text","path":"src/utils.py","file_url":"https://github.com/StigLidu/AdaExplore/blob/HEAD/src/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"20ac9f8ab3912be1"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.LG","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}