{"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/vision-language-models-are-zero-shot-reward","title":"Vision-Language Models are Zero-Shot Reward Models for Reinforcement Learning","arxiv_id":"2310.12921","date":"2023-10-19","proceeding":null,"authors":["Juan Rocamonde","Victoriano Montesinos","Elvis Nava","Ethan Perez","David Lindner"],"abstract":"Reinforcement learning (RL) requires either manually specifying a reward function, which is often infeasible, or learning a reward model from a large amount of human feedback, which is often very expensive. We study a more sample-efficient alternative: using pretrained vision-language models (VLMs) as zero-shot reward models (RMs) to specify tasks via natural language. We propose a natural and general approach to using VLMs as reward models, which we call VLM-RMs. We use VLM-RMs based on CLIP to train a MuJoCo humanoid to learn complex tasks without a manually specified reward function, such as kneeling, doing the splits, and sitting in a lotus position. For each of these tasks, we only provide a single sentence text prompt describing the desired task with minimal prompt engineering. We provide videos of the trained agents at: https://sites.google.com/view/vlm-rm. We can improve performance by providing a second \"baseline\" prompt and projecting out parts of the CLIP embedding space irrelevant to distinguish between goal and baseline. Further, we find a strong scaling effect for VLM-RMs: larger VLMs trained with more compute and data are better reward models. The failure modes of VLM-RMs we encountered are all related to known capability limitations of current VLMs, such as limited spatial reasoning ability or visually unrealistic environments that are far off-distribution for the VLM. We find that VLM-RMs are remarkably robust as long as the VLM is large enough. This suggests that future VLMs will become more and more useful reward models for a wide range of RL applications.","url_abs":"https://arxiv.org/abs/2310.12921v2","url_pdf":"https://arxiv.org/pdf/2310.12921v2.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":"vision-language-models-are-zero-shot-reward","repo_url":"https://github.com/alignmentresearch/vlmrm","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"mujoco","task_name":"MuJoCo"},{"task_slug":"prompt-engineering","task_name":"Prompt Engineering"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"sentence","task_name":"Sentence"},{"task_slug":"spatial-reasoning","task_name":"Spatial Reasoning"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[{"method_slug":"clip","method_name":"CLIP"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2310.12921","atlas_url":"https://app.syntology.ai/?focus=2310.12921","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2310.12921"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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/alignmentresearch/vlmrm","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":3,"unverified":2},"by_repo_kind":{"official":{"samples":5,"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":"356ec68c216ea0ca","entry":"get_clip_rewarded_env_name","repo":"alignmentresearch/vlmrm","repo_kind":"official","path":"src/vlmrm/envs/base.py","file_url":"https://github.com/alignmentresearch/vlmrm/blob/HEAD/src/vlmrm/envs/base.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"356ec68c216ea0ca"}},{"code_sha256_prefix":"c8660f2d2d63759f","entry":"get_make_env","repo":"alignmentresearch/vlmrm","repo_kind":"official","path":"src/vlmrm/envs/base.py","file_url":"https://github.com/alignmentresearch/vlmrm/blob/HEAD/src/vlmrm/envs/base.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":"c8660f2d2d63759f"}},{"code_sha256_prefix":"4111244697816bd1","entry":"is_3d_env","repo":"alignmentresearch/vlmrm","repo_kind":"official","path":"src/vlmrm/envs/base.py","file_url":"https://github.com/alignmentresearch/vlmrm/blob/HEAD/src/vlmrm/envs/base.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":"4111244697816bd1"}},{"code_sha256_prefix":"d7c6fc419a1cfd98","entry":"spawn","repo":"alignmentresearch/vlmrm","repo_kind":"official","path":"src/vlmrm/multiprocessing.py","file_url":"https://github.com/alignmentresearch/vlmrm/blob/HEAD/src/vlmrm/multiprocessing.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":"d7c6fc419a1cfd98"}},{"code_sha256_prefix":"9163212bb1ee550f","entry":"start_processes","repo":"alignmentresearch/vlmrm","repo_kind":"official","path":"src/vlmrm/multiprocessing.py","file_url":"https://github.com/alignmentresearch/vlmrm/blob/HEAD/src/vlmrm/multiprocessing.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":"9163212bb1ee550f"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}