# Learning Abstractions for Hierarchical Planning in Program-Synthesis Agents

arXiv: https://arxiv.org/abs/2602.00929 · HTML: https://mlsanity.com/paper/2602.00929/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Reinforcement learning for LLM agents, one of 16 directions in LLM agents and tool use.
- Builds on 39 of its 68 references that are in the map.
- First posted on arXiv 31 Jan 2026.

## Placed in

- LLM agents and tool use / [Reinforcement learning for LLM agents](https://mlsanity.com/agents/reinforcement-learning-for-llm-agents/)

## Builds on (most cited first, 30 of 39)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Tree of Thoughts: Deliberate Problem Solving with Large Language Models](https://mlsanity.com/paper/2305.10601/) (2023)
- [Voyager: An Open-Ended Embodied Agent with Large Language Models](https://mlsanity.com/paper/2305.16291/) (2023)
- [Dream to Control: Learning Behaviors by Latent Imagination](https://mlsanity.com/paper/1912.01603/) (2019)
- [Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model](https://mlsanity.com/paper/1911.08265/) (2019)
- [Mastering Diverse Domains through World Models](https://mlsanity.com/paper/2301.04104/) (2023)
- [Reasoning with Language Model is Planning with World Model](https://mlsanity.com/paper/2305.14992/) (2023)
- [Recurrent World Models Facilitate Policy Evolution](https://mlsanity.com/paper/1809.01999/) (2018)
- [Visual Programming: Compositional visual reasoning without training](https://mlsanity.com/paper/2211.11559/) (2022)
- [LLM+P: Empowering Large Language Models with Optimal Planning Proficiency](https://mlsanity.com/paper/2304.11477/) (2023)
- [The Option-Critic Architecture](https://mlsanity.com/paper/1609.05140/) (2016)
- [FeUdal Networks for Hierarchical Reinforcement Learning](https://mlsanity.com/paper/1703.01161/) (2017)
- [Mastering Atari Games with Limited Data](https://mlsanity.com/paper/2111.00210/) (2021)
- [Large Language Models as Commonsense Knowledge for Large-Scale Task Planning](https://mlsanity.com/paper/2305.14078/) (2023)
- [Leveraging Pre-trained Large Language Models to Construct and Utilize World Models for Model-based Task Planning](https://mlsanity.com/paper/2305.14909/) (2023)
- [BALROG: Benchmarking Agentic LLM and VLM Reasoning On Games](https://mlsanity.com/paper/2411.13543/) (2024)
- [Imagination-Augmented Agents for Deep Reinforcement Learning](https://mlsanity.com/paper/1707.06203/) (2017)
- [WorldCoder, a Model-Based LLM Agent: Building World Models by Writing Code and Interacting with the Environment](https://mlsanity.com/paper/2402.12275/) (2024)
- [MiniHack the Planet: A Sandbox for Open-Ended Reinforcement Learning Research](https://mlsanity.com/paper/2109.13202/) (2021)
- [Don't Overthink it. Preferring Shorter Thinking Chains for Improved LLM Reasoning](https://mlsanity.com/paper/2505.17813/) (2025)
- [On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models](https://mlsanity.com/paper/1511.09249/) (2015)
- [Reason for Future, Act for Now: A Principled Framework for Autonomous LLM Agents with Provable Sample Efficiency](https://mlsanity.com/paper/2309.17382/) (2023)
- [Human-Level Reinforcement Learning through Theory-Based Modeling, Exploration, and Planning](https://mlsanity.com/paper/2107.12544/) (2021)
- [Large Language Models for Planning: A Comprehensive and Systematic Survey](https://mlsanity.com/paper/2505.19683/) (2025)
- [VisualPredicator: Learning Abstract World Models with Neuro-Symbolic Predicates for Robot Planning](https://mlsanity.com/paper/2410.23156/) (2024)
- [Learning adaptive planning representations with natural language guidance](https://mlsanity.com/paper/2312.08566/) (2023)
- [LMAct: A Benchmark for In-Context Imitation Learning with Long Multimodal Demonstrations](https://mlsanity.com/paper/2412.01441/) (2024)
- [PoE-World: Compositional World Modeling with Products of Programmatic Experts](https://mlsanity.com/paper/2505.10819/) (2025)

## Built on by (most cited first, 3 of 3)

- [Tycho: Active Abstraction with Programmatic World Models for ARC-AGI-3](https://mlsanity.com/paper/2607.28287/) (2026)
- [Code to Control: Synthesizing Parameterized Reactive Controllers](https://mlsanity.com/paper/2609.38733/) (2026)
- [Schema: Discovering Unknown Environments via Agentic Program Induction](https://mlsanity.com/paper/2609.39140/) (2026)

## Abstract (arXiv)

Humans learn abstractions and use them to plan efficiently to quickly generalize across tasks -- an ability that remains challenging for state-of-the-art large language model (LLM) agents and deep reinforcement learning (RL) systems. Inspired by the cognitive science of how people form abstractions and intuitive theories of their world knowledge, Theory-Based RL (TBRL) systems, such as TheoryCoder, exhibit strong generalization through effective use of abstractions. However, they heavily rely on human-provided abstractions and sidestep the abstraction-learning problem. We introduce TheoryCoder-2, a new TBRL agent that leverages LLMs' in-context learning ability to actively learn reusable abstractions rather than relying on hand-specified ones, by synthesizing abstractions from experience and integrating them into a hierarchical planning process. We conduct experiments on diverse environments, including BabyAI, Minihack and VGDL games like Sokoban. We find that TheoryCoder-2 is significantly more sample-efficient than baseline LLM agents augmented with classical planning domain construction, reasoning-based planning, and prior program-synthesis agents such as WorldCoder. TheoryCoder-2 is able to solve complex tasks that the baselines fail, while only requiring minimal human prompts, unlike prior TBRL systems.