# GRAIL: Autonomous Concept Grounding for Neuro-Symbolic Reinforcement Learning

arXiv: https://arxiv.org/abs/2604.16871 · HTML: https://mlsanity.com/paper/2604.16871/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reinforcement learning for reasoning, one of 17 directions in LLM reasoning.
- Builds on 62 of its 81 references that are in the map.
- First posted on arXiv 18 Apr 2026.

## Placed in

- LLM reasoning / [Reinforcement learning for reasoning](https://mlsanity.com/lr/reinforcement-learning-for-reasoning/)
- Deep reinforcement learning methods / [Interpretable and logic-guided RL](https://mlsanity.com/deeprl/interpretable-and-logic-guided-rl/)

## Builds on (most cited first, 30 of 62)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [You Only Look Once: Unified, Real-Time Object Detection](https://mlsanity.com/paper/1506.02640/) (2015)
- [Playing Atari with Deep Reinforcement Learning](https://mlsanity.com/paper/1312.5602/) (2013)
- [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://mlsanity.com/paper/2106.01345/) (2021)
- [Concept Bottleneck Models](https://mlsanity.com/paper/2007.04612/) (2020)
- [Object-Centric Learning with Slot Attention](https://mlsanity.com/paper/2006.15055/) (2020)
- [Rainbow: Combining Improvements in Deep Reinforcement Learning](https://mlsanity.com/paper/1710.02298/) (2017)
- [The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences From Natural Supervision](https://mlsanity.com/paper/1904.12584/) (2019)
- [Agent57: Outperforming the Atari Human Benchmark](https://mlsanity.com/paper/2003.13350/) (2020)
- [MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learning](https://mlsanity.com/paper/2109.12674/) (2021)
- [Fast Segment Anything](https://mlsanity.com/paper/2306.12156/) (2023)
- [Stabilizing Transformers for Reinforcement Learning](https://mlsanity.com/paper/1910.06764/) (2019)
- [Conditional Object-Centric Learning from Video](https://mlsanity.com/paper/2111.12594/) (2021)
- [Habitat 3.0: A Co-Habitat for Humans, Avatars and Robots](https://mlsanity.com/paper/2310.13724/) (2023)
- [Learning Explanatory Rules from Noisy Data](https://mlsanity.com/paper/1711.04574/) (2017)
- [Programmatically Interpretable Reinforcement Learning](https://mlsanity.com/paper/1804.02477/) (2018)
- [SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition](https://mlsanity.com/paper/2001.02407/) (2020)
- [CrossQ: Batch Normalization in Deep Reinforcement Learning for Greater Sample Efficiency and Simplicity](https://mlsanity.com/paper/1902.05605/) (2019)
- [Right for the Right Concept: Revising Neuro-Symbolic Concepts by Interacting with their Explanations](https://mlsanity.com/paper/2011.12854/) (2020)
- [SDRL: Interpretable and Data-efficient Deep Reinforcement Learning Leveraging Symbolic Planning](https://mlsanity.com/paper/1811.00090/) (2018)
- [Deep Differentiable Logic Gate Networks](https://mlsanity.com/paper/2210.08277/) (2022)
- [Combining Reinforcement Learning and Constraint Programming for Combinatorial Optimization](https://mlsanity.com/paper/2006.01610/) (2020)
- [Neural Logic Reinforcement Learning](https://mlsanity.com/paper/1904.10729/) (2019)
- [Predicate Invention for Bilevel Planning](https://mlsanity.com/paper/2203.09634/) (2022)
- [An Investigation into Pre-Training Object-Centric Representations for Reinforcement Learning](https://mlsanity.com/paper/2302.04419/) (2023)
- [Interpretable and Explainable Logical Policies via Neurally Guided Symbolic Abstraction](https://mlsanity.com/paper/2306.01439/) (2023)
- [Self-supervised Visual Reinforcement Learning with Object-centric Representations](https://mlsanity.com/paper/2011.14381/) (2020)
- [Explainable Deep Reinforcement Learning: State of the Art and Challenges](https://mlsanity.com/paper/2301.09937/) (2023)
- [What's Left? Concept Grounding with Logic-Enhanced Foundation Models](https://mlsanity.com/paper/2310.16035/) (2023)
- [Neuro-Symbolic Reinforcement Learning with First-Order Logic](https://mlsanity.com/paper/2110.10963/) (2021)

## Abstract (arXiv)

Neuro-symbolic Reinforcement Learning (NeSy-RL) combines symbolic reasoning with gradient-based optimization to achieve interpretable and generalizable policies. Relational concepts, such as "left of" or "close by", serve as foundational building blocks that structure how agents perceive and act. However, conventional approaches require human experts to manually define these concepts, limiting adaptability since concept semantics vary across environments. We propose GRAIL (Grounding Relational Agents through Interactive Learning), a framework that autonomously grounds relational concepts through environmental interaction. GRAIL leverages large language models (LLMs) to provide generic concept representations as weak supervision, then refines them to capture environment-specific semantics. This approach addresses both sparse reward signals and concept misalignment prevalent in underdetermined environments. Experiments on the Atari games Kangaroo, Seaquest, and Skiing demonstrate that GRAIL matches or outperforms agents with manually crafted concepts in simplified settings, and reveals informative trade-offs between reward maximization and high-level goal completion in the full environment.