Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.20459 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916712428863488 |
|---|---|
| author | Amor, Heni Ben Graesser, Laura Iscen, Atil D'Ambrosio, David Abeyruwan, Saminda Bewley, Alex Zhou, Yifan Kalirathinam, Kamalesh Mishra, Swaroop Sanketi, Pannag |
| author_facet | Amor, Heni Ben Graesser, Laura Iscen, Atil D'Ambrosio, David Abeyruwan, Saminda Bewley, Alex Zhou, Yifan Kalirathinam, Kamalesh Mishra, Swaroop Sanketi, Pannag |
| contents | We demonstrate the ability of large language models (LLMs) to perform iterative self-improvement of robot policies. An important insight of this paper is that LLMs have a built-in ability to perform (stochastic) numerical optimization and that this property can be leveraged for explainable robot policy search. Based on this insight, we introduce the SAS Prompt (Summarize, Analyze, Synthesize) -- a single prompt that enables iterative learning and adaptation of robot behavior by combining the LLM's ability to retrieve, reason and optimize over previous robot traces in order to synthesize new, unseen behavior. Our approach can be regarded as an early example of a new family of explainable policy search methods that are entirely implemented within an LLM. We evaluate our approach both in simulation and on a real-robot table tennis task. Project website: sites.google.com/asu.edu/sas-llm/ |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_20459 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | SAS-Prompt: Large Language Models as Numerical Optimizers for Robot Self-Improvement Amor, Heni Ben Graesser, Laura Iscen, Atil D'Ambrosio, David Abeyruwan, Saminda Bewley, Alex Zhou, Yifan Kalirathinam, Kamalesh Mishra, Swaroop Sanketi, Pannag Robotics We demonstrate the ability of large language models (LLMs) to perform iterative self-improvement of robot policies. An important insight of this paper is that LLMs have a built-in ability to perform (stochastic) numerical optimization and that this property can be leveraged for explainable robot policy search. Based on this insight, we introduce the SAS Prompt (Summarize, Analyze, Synthesize) -- a single prompt that enables iterative learning and adaptation of robot behavior by combining the LLM's ability to retrieve, reason and optimize over previous robot traces in order to synthesize new, unseen behavior. Our approach can be regarded as an early example of a new family of explainable policy search methods that are entirely implemented within an LLM. We evaluate our approach both in simulation and on a real-robot table tennis task. Project website: sites.google.com/asu.edu/sas-llm/ |
| title | SAS-Prompt: Large Language Models as Numerical Optimizers for Robot Self-Improvement |
| topic | Robotics |
| url | https://arxiv.org/abs/2504.20459 |