mlsanityData to 7 Oct 2026

Can LLMs Translate Human Instructions into a Reinforcement Learning Agent's Internal Emergent Symbolic Representation?

arXiv 2510.24259 ↗First posted 28 Oct 2025cs.CL cs.ROcited by 0

On the map: LLM agents and tool use → Reinforcement learning for LLM agents

Abstract (arXiv)

Emergent symbolic representations are critical for enabling developmental learning agents to plan and generalize across tasks. In this work, we investigate whether large language models (LLMs) can translate human natural language instructions into the internal symbolic representations that emerge during hierarchical reinforcement learning. We apply a structured evaluation framework to measure the translation performance of commonly seen LLMs -- GPT, Claude, Deepseek and Grok -- across different internal symbolic partitions generated by a hierarchical reinforcement learning algorithm in the Ant Maze and Ant Fall environments. Our findings reveal that although LLMs demonstrate some ability to translate natural language into a symbolic representation of the environment dynamics, their performance is highly sensitive to partition granularity and task complexity. The results expose limitations in current LLMs capacity for representation alignment, highlighting the need for further research on robust alignment between language and internal agent representations.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reinforcement learning for LLM agents, one of 16 directions in LLM agents and tool use. Builds on 9 of its 17 references that are in the map. First posted on arXiv 28 Oct 2025.

Builds on

9 of 17 references.

References · 9

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,155
  3. 2016Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translationcited by 1,471
  4. 2023A Comprehensive Overview of Large Language Modelscited by 639
  5. 2018Data-Efficient Hierarchical Reinforcement Learningcited by 436
  6. 2016Benchmarking Deep Reinforcement Learning for Continuous Controlcited by 361
  7. 2018Symbol Emergence in Cognitive Developmental Systems: a Surveycited by 39
  8. 2023Goal Space Abstraction in Hierarchical Reinforcement Learning via Set-Based Reachability Analysiscited by 6
  9. 2024Reconciling Spatial and Temporal Abstractions for Goal Representationcited by 6

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Understanding Large Language Models
  2. 2024Mental Modeling of Reinforcement Learning Agents by Language Models
  3. 2023Synergistic Integration of Large Language Models and Cognitive Architectures for Robust AI: An Exploratory Analysis
  4. 2025From Word to World: Can Large Language Models be Implicit Text-based World Models?
  5. 2025Why do AI agents communicate in human language?
  6. 2026Exploring Collaboration between a language and a non-language agent
  7. 2026Rethinking Agentic Reinforcement Learning In Large Language Models
  8. 2024Reinforcement Learning for Aligning Large Language Models Agents with Interactive Environments: Quantifying and Mitigating Prompt Overfitting