The DOI matches its Crossref record; 60 of the 60 references it cites were found in a registry or an index.
Code as Policies: Language Model Programs for Embodied Control
Open accessAbstract
Large language models (LLMs) trained on code-completion have been shown to be capable of synthesizing simple Python programs from docstrings [1]. We find that these code-writing LLMs can be re-purposed to write robot policy code, given natural language commands. Specifically, policy code can express functions or feedback loops that process perception outputs (e.g., from object detectors [2], [3]) and parameterize control primitive APIs. When provided as input several example language commands (formatted as comments) followed by corresponding policy code (via few-shot prompting), LLMs can take in new commands and autonomously re-compose API calls to generate new policy code respectively. By chaining classic logic structures and referencing third-party libraries (e.g., NumPy, Shapely) to perform arithmetic, LLMs used in this way can write robot policies that (i) exhibit spatial-geometric reasoning, (ii) generalize to new instructions, and (iii) prescribe precise values (e.g., velocities) to ambiguous descriptions (‘faster’) depending on context (i.e., behavioral commonsense). This paper presents Code as Policies: a robot-centric formulation of language model generated programs (LMPs) that can represent reactive policies (e.g., impedance controllers), as well as waypoint-based policies (vision-based pick and place, trajectory-based control), demonstrated across multiple real robot platforms. Central to our approach is prompting hierarchical code-gen (recursively defining undefined functions), which can write more complex code and also improves state-of-the-art to solve 39.8% of problems on the HumanEval [1] benchmark. Code and videos are available at https://code-as-policies.github.io
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“LLM-generated robotic task plans are highly sensitive to prompt phrasing (Liang et al., 2023) and vulnerable to adversarial manipulation (Zhang et al., 2024c).”
From Large Language Model Reasoning Failures (2026) · cites this paper as Liang et al., 2023
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Naive LLMs and VLMs generate plans step by step, lacking mechanisms to detect and correct earlier mistakes or execution errors (Liang et al., 2023; Huang et al., 2022b; Duan et al., 2024).”
From Large Language Model Reasoning Failures (2026) · cites this paper as Liang et al., 2023
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Incorporating feedback mechanisms or explicit error-handling strategies significantly reduces these errors (Liang et al., 2023; Wang et al., 2023a).”
From Large Language Model Reasoning Failures (2026) · cites this paper as Liang et al., 2023
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“In personalization and embodied control, agents author and revise executable policies, simulators, and skill libraries in response to environment feedback [9, 10, 32].”
From Code as Agent Harness (2026) · cites this paper as [10]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Programs are increasingly used as the medium through which LLM agents reason, act, and model their environments.”
From Code as Agent Harness (2026) · cites this paper as [10]
Experimental — accuracy not yet measured
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Code and classical robotics software define the admissibility boundary by exposing typed robot APIs, parameterizing primitive skills, calling geometric libraries, invoking motion planners, and supporting inspection, replay, versioning, and verification [124, 10, 434, 435].”
From Code as Agent Harness (2026) · cites this paper as [10]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Whereas previous work needed to compose multiple models to this end (Ahn et al., 2022; Kwon et al., 2024; Liang et al., 2023; Vemprala et al., 2023)”
From Gemini Robotics: Bringing AI into the Physical World (2025) · cites this paper as Liang et al., 2023
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Trained on large datasets of images and language, VLMs can provide powerful visual semantic understanding and useful guidance to robotic tasks, such as which object a manipulator should pick up or which goal a mobile robot should reach [8, 9, 10].”
From RoboPoint: A Vision-Language Model for Spatial Affordance Prediction for Robotics (2024) · cites this paper as [9]
The cited source says what the claim attributes to it.
In the citing paper
“Several works [8, 9, 10] have shown that language model are capable planners for robotic tasks.”
From RoboPoint: A Vision-Language Model for Spatial Affordance Prediction for Robotics (2024) · cites this paper as [9]
The cited source says what the claim attributes to it.
In the citing paper
“We opt to use code as the action space instead of low-level motor commands because programs can naturally represent temporally extended and compositional actions [16, 22], which are essential for many long-horizon tasks in Minecraft.”
From Voyager: An Open-Ended Embodied Agent with Large Language Models (2023) · cites this paper as [16]
The cited source says what the claim attributes to it.
In the citing paper
“robotic and embodied agents use generated programs as executable policies for interacting with physical or simulated worlds [9, 10]”
From Code as Agent Harness (2026) · cites this paper as [10]
The cited source says what the claim attributes to it.
In the citing paper
“Recent advances in large language model (LLM) based agents harness the world knowledge encapsulated in pre-trained LLMs to generate consistent action plans or executable policies [16, 22, 19].”
From Voyager: An Open-Ended Embodied Agent with Large Language Models (2023) · cites this paper as [16]
The cited source says what the claim attributes to it.
In the citing paper
“As a grounding interface, it translates high-level intent from LLMs into embodiment-respecting commands through primitive skill calls [9, 110, 111, 112], synthesized Python control policies [10, 33, 124, 117, 118], and structured behavior-tree programs [34].”
From Code as Agent Harness (2026) · cites this paper as [10]
The cited source says what the claim attributes to it.
In the citing paper
“Many existing approaches leveraging foundation models include CodeAsPolicies [6], Voyager [35], VIMA [92], VoxPoser [77].”
From Embodied Agent Interface: Benchmarking LLMs for Embodied Decision Making (2024) · cites this paper as [6]
The check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“there has been a performance revolution of sorts there (Yao et al., 2023b;Liang et al., 2023;Wang et al., 2023b).”
From LLMs Can't Plan, But Can Help Planning in LLM-Modulo Frameworks (2024) · cites this paper as Liang et al., 2023
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372
Chain-Of-Thought Prompting Elicits Reasoning in Large Language Models
Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E., Le, Q. V., & Zhou, D. (2022) · Advances in Neural Information Processing Systems 35 · doi:10.52202/068431-1800 · Cited by 4,505
Self-citationTraining language models to follow instructions with human feedback
Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C. L., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., et al. (2022) · arXiv · doi:10.48550/arxiv.2203.02155 · Cited by 4,208
PaLM: Scaling Language Modeling with Pathways
Chowdhery, A., Narang, S., Devlin, J., Bosma, M., Mishra, G., Roberts, A., Barham, P., Chung, H. W., Sutton, C., Gehrmann, S., et al. (2022) · arXiv · doi:10.48550/arxiv.2204.02311 · Cited by 2,590
Large language models are zero-shot reasoners
Kojima, T., Gu, S. (., Reid, M., Matsuo, Y., & Iwasawa, Y. (2022) · Advances in Neural Information Processing Systems 35 · doi:10.52202/068431-1613 · Cited by 2,013
Open-vocabulary Object Detection via Vision and Language Knowledge DistillationPreprint
Gu, X.-Y., Lin, T.-Y., Kuo, W.-C., & Cui, Y. (2021) · International Conference on Learning Representations · doi:10.48550/arxiv.2104.13921 · Cited by 1,464
Evaluating Large Language Models Trained on Code
Chen, M., Tworek, J., Jun, H., Yuan, Q., Pinto, H. P. D. O., Kaplan, J., Edwards, H., Burda, Y., Joseph, N., Brockman, G., et al. (2021) · arXiv · doi:10.48550/arxiv.2107.03374 · Cited by 1,436
Understanding Natural Language Commands for Robotic Navigation and Mobile Manipulation
Tellex, S., Kollar, T., Dickerson, S., Walter, M. R., Banerjee, A., Teller, S. J., & Roy, N. (2011) · AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v25i1.7979 · Cited by 903
MDETR - Modulated Detection for End-to-End Multi-Modal Understanding
Kamath, A., Singh, M., LeCun, Y., Synnaeve, G., Misra, I., & Carion, N. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00180 · Cited by 802
Do As I Can, Not As I Say: Grounding Language in Robotic Affordances
Ahn, M., Brohan, A., Brown, N., Chebotar, Y., Cortes, O., David, B., Finn, C., Fu, C., Gopalakrishnan, K., Hausman, K., et al. (2022) · arXiv · doi:10.48550/arxiv.2204.01691 · Cited by 624
Self-citation
Research works printed but not linked: 2
- cideron, 2019, Self-educated language agent with hindsight experience replay for instruction following, DeepMind
- kalashnikov, 2018, Scalable deep reinforcement learning for vision-based robotic manipulation, CoRL
Cited by
Sorted by citations
The Rise and Potential of Large Language Model Based Agents: A Survey
Xi, Z., Chen, W., Guo, X., He, W., Ding, Y., Hong, B., Zhang, M., Wang, J., Jin, S., Zhou, E., et al. (2023) · arXiv · doi:10.48550/arxiv.2309.07864 · Cited by 972
The rise and potential of large language model based agents: a survey
Xi, Z., Chen, W., Guo, X., He, W., Ding, Y., Hong, B., Zhang, M., Wang, J., Jin, S., Zhou, E., et al. (2025) · Science China Information Sciences · doi:10.1007/s11432-024-4222-0 · Cited by 970
(2022) · doi:10.15607/rss.2023.xix.025 · Cited by 916
Self-citationChatGPT for Robotics: Design Principles and Model Abilities
Vemprala, S. H., Bonatti, R., Bucker, A., & Kapoor, A. (2024) · IEEE Access · doi:10.1109/access.2024.3387941 · Cited by 570
DriveGPT4: Interpretable End-to-End Autonomous Driving Via Large Language Model
Xu, Z., Zhang, Y., Xie, E., Zhen, Z., Guo, Y. F., Wong, K. K., Li, Z., & Zhao, H. (2024) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2024.3440097 · Cited by 640
Metadata
- Authors
- Liang, J., Huang, W., Xia, F., Xu, P., Hausman, K., Ichter, B., Florence, P., & Zeng, A.
- Venue
- 2023 IEEE International Conference on Robotics and Automation (ICRA)
- Publisher
- IEEE
- Published
- 2023-05-29
- Pages
- 9493-9500
- ISSN
- 1050-4729 (ISSN-L)
- OpenAlex
- W4383097638
- Cited by
- 767
- References
- 59
- Citations by year
- 2026: 231 · 2025: 286 · 2024: 201 · 2023: 48
- FWCI
- 327.59
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- Proceedings - IEEE International Conference on Robotics and Automation/Proceedings (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Engineering › Control and Systems Engineering › Robot Manipulation and Learning
- Physical Sciences › Computer Science › Artificial Intelligence › AI-based Problem Solving and Planning
- Physical Sciences › Computer Science › Information Systems › Software Engineering Research
- Keywords
- large language models, code generation, robot policy, embodied control, natural language commands, few-shot prompting, trajectory-based control, impedance control
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
