# Sequential Planning in Large Partially Observable Environments guided by LLMs

arXiv: https://arxiv.org/abs/2312.07368 · HTML: https://mlsanity.com/paper/2312.07368/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 7 of its 13 references that are in the map.
- First posted on arXiv 12 Dec 2023.

## Builds on (most cited first, 7 of 7)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Do As I Can, Not As I Say: Grounding Language in Robotic Affordances](https://mlsanity.com/paper/2204.01691/) (2022)
- [ScienceWorld: Is your Agent Smarter than a 5th Grader?](https://mlsanity.com/paper/2203.07540/) (2022)
- [Keep CALM and Explore: Language Models for Action Generation in Text-based Games](https://mlsanity.com/paper/2010.02903/) (2020)
- [CLIN: A Continually Learning Language Agent for Rapid Task Adaptation and Generalization](https://mlsanity.com/paper/2310.10134/) (2023)
- [Deep Reinforcement Learning with a Natural Language Action Space](https://mlsanity.com/paper/1511.04636/) (2015)
- [Graph Constrained Reinforcement Learning for Natural Language Action Spaces](https://mlsanity.com/paper/2001.08837/) (2020)

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

- [GAAMA: Graph Augmented Associative Memory for Agents](https://mlsanity.com/paper/2603.27910/) (2026)
- [Disentangling Exploration of Large Language Models by Optimal Exploitation](https://mlsanity.com/paper/2501.08925/) (2025)
- [CoSPlan: Corrective Sequential Planning via Scene Graph Incremental Updates](https://mlsanity.com/paper/2512.10342/) (2025)

## Abstract (arXiv)

Sequential planning in large state space and action space quickly becomes intractable due to combinatorial explosion of the search space. Heuristic methods, like monte-carlo tree search, though effective for large state space, but struggle if action space is large. Pure reinforcement learning methods, relying only on reward signals, needs prohibitively large interactions with the environment to device a viable plan. If the state space, observations and actions can be represented in natural language then Large Language models (LLM) can be used to generate action plans. Recently several such goal-directed agents like Reflexion, CLIN, SayCan were able to surpass the performance of other state-of-the-art methods with minimum or no task specific training. But they still struggle with exploration and get stuck in local optima. Their planning capabilities are limited by the limited reasoning capability of the foundational LLMs on text data. We propose a hybrid agent "neoplanner", that synergizes both state space search with queries to foundational LLM to get the best action plan. The reward signals are quantitatively used to drive the search. A balance of exploration and exploitation is maintained by maximizing upper confidence bounds of values of states. In places where random exploration is needed, the LLM is queried to generate an action plan. Learnings from each trial are stored as entity relationships in text format. Those are used in future queries to the LLM for continual improvement. Experiments in the Scienceworld environment reveals a 124% improvement from the current best method in terms of average reward gained across multiple tasks.