{"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/self-play-with-execution-feedback-improving","title":"Self-play with Execution Feedback: Improving Instruction-following Capabilities of Large Language Models","arxiv_id":"2406.13542","date":"2024-06-19","proceeding":null,"authors":["Guanting Dong","Keming Lu","Chengpeng Li","Tingyu Xia","Bowen Yu","Chang Zhou","Jingren Zhou"],"abstract":"One core capability of large language models (LLMs) is to follow natural language instructions. However, the issue of automatically constructing high-quality training data to enhance the complex instruction-following abilities of LLMs without manual annotation remains unresolved. In this paper, we introduce AutoIF, the first scalable and reliable method for automatically generating instruction-following training data. AutoIF transforms the validation of instruction-following data quality into code verification, requiring LLMs to generate instructions, the corresponding code to check the correctness of the instruction responses, and unit test samples to verify the code's correctness. Then, execution feedback-based rejection sampling can generate data for Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF) training. AutoIF achieves significant improvements across three training algorithms, SFT, Offline DPO, and Online DPO, when applied to the top open-source LLMs, Qwen2 and LLaMA3, in self-alignment and strong-to-weak distillation settings. Our code is publicly available at https://github.com/QwenLM/AutoIF.","url_abs":"https://arxiv.org/abs/2406.13542v3","url_pdf":"https://arxiv.org/pdf/2406.13542v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"self-play-with-execution-feedback-improving","repo_url":"https://github.com/QwenLM/AutoIF","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"instruction-following","task_name":"Instruction Following"}],"methods":[{"method_slug":"dpo","method_name":"DPO"},{"method_slug":"sft","method_name":"SFT"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/instruction-following-on-ifeval","task":"Instruction Following","dataset":"IFEval","model":"AutoIF (Llama3 70B)","rank_in_archive_order":1,"of":4,"metrics":{"Inst-level loose-accuracy":"90.4","Inst-level strict-accuracy":"86.7","Prompt-level loose-accuracy":"85.6","Prompt-level strict-accuracy":"80.2"},"uses_additional_data":false},{"leaderboard":"/sota/instruction-following-on-ifeval","task":"Instruction Following","dataset":"IFEval","model":"AutoIF (Qwen2 72B)","rank_in_archive_order":2,"of":4,"metrics":{"Inst-level loose-accuracy":"88","Inst-level strict-accuracy":"86.1","Prompt-level loose-accuracy":"82.3","Prompt-level strict-accuracy":"80.2"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2406.13542","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2406.13542"}},"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":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/QwenLM/AutoIF","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran":1},"by_repo_kind":{"official":{"samples":1,"ran":1,"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":"0c720b590a991017","entry":"random_sample","repo":"QwenLM/AutoIF","repo_kind":"official","path":"code_dpo/2_dpo_data_query_construct.py","file_url":"https://github.com/QwenLM/AutoIF/blob/HEAD/code_dpo/2_dpo_data_query_construct.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"0c720b590a991017"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}