{"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/large-language-models-can-self-improve-at-web","title":"Large Language Models Can Self-Improve At Web Agent Tasks","arxiv_id":"2405.20309","date":"2024-05-30","proceeding":null,"authors":["Ajay Patel","Markus Hofmarcher","Claudiu Leoveanu-Condrei","Marius-Constantin Dinu","Chris Callison-Burch","Sepp Hochreiter"],"abstract":"Training models to act as agents that can effectively navigate and perform actions in a complex environment, such as a web browser, has typically been challenging due to lack of training data. Large language models (LLMs) have recently demonstrated some capability to navigate novel environments as agents in a zero-shot or few-shot fashion, purely guided by natural language instructions as prompts. Recent research has also demonstrated LLMs have the capability to exceed their base performance through self-improvement, i.e. fine-tuning on data generated by the model itself. In this work, we explore the extent to which LLMs can self-improve their performance as agents in long-horizon tasks in a complex environment using the WebArena benchmark. In WebArena, an agent must autonomously navigate and perform actions on web pages to achieve a specified objective. We explore fine-tuning on three distinct synthetic training data mixtures and achieve a 31\\% improvement in task completion rate over the base model on the WebArena benchmark through a self-improvement procedure. We additionally contribute novel evaluation metrics for assessing the performance, robustness, capabilities, and quality of trajectories of our fine-tuned agent models to a greater degree than simple, aggregate-level benchmark scores currently used to measure self-improvement.","url_abs":"https://arxiv.org/abs/2405.20309v2","url_pdf":"https://arxiv.org/pdf/2405.20309v2.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":"large-language-models-can-self-improve-at-web","repo_url":"https://github.com/AjayP13/webdreamer","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"navigate","task_name":"Navigate"}],"methods":[{"method_slug":"base","method_name":"BASE"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2405.20309","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2405.20309"}},"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/AjayP13/webdreamer","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":3,"unverified":5},"by_repo_kind":{"official":{"samples":8,"ran":3,"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":"243dd0dde1f8275f","entry":"get_jax_device","repo":"AjayP13/webdreamer","repo_kind":"official","path":"src/project/devices.py","file_url":"https://github.com/AjayP13/webdreamer/blob/HEAD/src/project/devices.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"243dd0dde1f8275f"}},{"code_sha256_prefix":"b3e7ae0bcff74b3f","entry":"get_tf_device","repo":"AjayP13/webdreamer","repo_kind":"official","path":"src/project/devices.py","file_url":"https://github.com/AjayP13/webdreamer/blob/HEAD/src/project/devices.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b3e7ae0bcff74b3f"}},{"code_sha256_prefix":"722a305d9794bd53","entry":"get_torch_device","repo":"AjayP13/webdreamer","repo_kind":"official","path":"src/project/devices.py","file_url":"https://github.com/AjayP13/webdreamer/blob/HEAD/src/project/devices.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"722a305d9794bd53"}},{"code_sha256_prefix":"c9b3c3d69b6ac466","entry":"call_server","repo":"AjayP13/webdreamer","repo_kind":"official","path":"src/tasks/webarena_eval/_llm_server.py","file_url":"https://github.com/AjayP13/webdreamer/blob/HEAD/src/tasks/webarena_eval/_llm_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":"c9b3c3d69b6ac466"}},{"code_sha256_prefix":"69c9d990fdc6e372","entry":"get_persistent_dir","repo":"AjayP13/webdreamer","repo_kind":"official","path":"src/project/persistent_storage.py","file_url":"https://github.com/AjayP13/webdreamer/blob/HEAD/src/project/persistent_storage.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":"69c9d990fdc6e372"}},{"code_sha256_prefix":"d6c69b8f8e7a68e9","entry":"run_cloudflared","repo":"AjayP13/webdreamer","repo_kind":"official","path":"src/project/serve.py","file_url":"https://github.com/AjayP13/webdreamer/blob/HEAD/src/project/serve.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":"d6c69b8f8e7a68e9"}},{"code_sha256_prefix":"8be329d70b8a59e5","entry":"run_jupyter","repo":"AjayP13/webdreamer","repo_kind":"official","path":"src/project/serve.py","file_url":"https://github.com/AjayP13/webdreamer/blob/HEAD/src/project/serve.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":"8be329d70b8a59e5"}},{"code_sha256_prefix":"d3305f44defae2cf","entry":"run_ngrok","repo":"AjayP13/webdreamer","repo_kind":"official","path":"src/project/serve.py","file_url":"https://github.com/AjayP13/webdreamer/blob/HEAD/src/project/serve.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":"d3305f44defae2cf"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}