# Emergence of Abstract State Representations in Embodied Sequence Modeling

arXiv: https://arxiv.org/abs/2311.02171 · HTML: https://mlsanity.com/paper/2311.02171/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 22 references that are in the map.
- First posted on arXiv 3 Nov 2023.

## Builds on (most cited first, 18 of 18)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Dream to Control: Learning Behaviors by Latent Imagination](https://mlsanity.com/paper/1912.01603/) (2019)
- [Multitask Prompted Training Enables Zero-Shot Task Generalization](https://mlsanity.com/paper/2110.08207/) (2021)
- [Solving Quantitative Reasoning Problems with Language Models](https://mlsanity.com/paper/2206.14858/) (2022)
- [Understanding intermediate layers using linear classifier probes](https://mlsanity.com/paper/1610.01644/) (2016)
- [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://mlsanity.com/paper/2106.02039/) (2021)
- [Designing and Interpreting Probes with Control Tasks](https://mlsanity.com/paper/1909.03368/) (2019)
- [What do you learn from context? Probing for sentence structure in contextualized word representations](https://mlsanity.com/paper/1905.06316/) (2019)
- [Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Task](https://mlsanity.com/paper/2210.13382/) (2022)
- [BabyAI: A Platform to Study the Sample Efficiency of Grounded Language Learning](https://mlsanity.com/paper/1810.08272/) (2018)
- [Pre-Trained Language Models for Interactive Decision-Making](https://mlsanity.com/paper/2202.01771/) (2022)
- [Episodic Transformer for Vision-and-Language Navigation](https://mlsanity.com/paper/2105.06453/) (2021)
- [Can Language Models Encode Perceptual Structure Without Grounding? A Case Study in Color](https://mlsanity.com/paper/2109.06129/) (2021)
- [Representation Matters: Offline Pretraining for Sequential Decision Making](https://mlsanity.com/paper/2102.05815/) (2021)
- [Masked Trajectory Models for Prediction, Representation, and Control](https://mlsanity.com/paper/2305.02968/) (2023)

## Built on by (most cited first, 9 of 9)

- [How Do Vision-Language Models Process Conflicting Information Across Modalities?](https://mlsanity.com/paper/2507.01790/) (2025)
- [What Do Language Models Hear? Probing for Auditory Representations in Language Models](https://mlsanity.com/paper/2402.16998/) (2024)
- [Do Music Generation Models Encode Music Theory?](https://mlsanity.com/paper/2410.00872/) (2024)
- [Revisiting the Othello World Model Hypothesis](https://mlsanity.com/paper/2503.04421/) (2025)
- [LLM-Based Offline Learning for Embodied Agents via Consistency-Guided Reward Ensemble](https://mlsanity.com/paper/2411.17135/) (2024)
- [Prevailing Research Areas for Music AI in the Era of Foundation Models](https://mlsanity.com/paper/2409.09378/) (2024)
- [Semantic Skill Grounding for Embodied Instruction-Following in Cross-Domain Environments](https://mlsanity.com/paper/2408.01024/) (2024)
- [What if Othello-Playing Language Models Could See?](https://mlsanity.com/paper/2507.14520/) (2025)
- [A Close Look At World Model Recovery In Supervised Fine-Tuned LLM Planners](https://mlsanity.com/paper/2606.03685/) (2026)

## Abstract (arXiv)

Decision making via sequence modeling aims to mimic the success of language models, where actions taken by an embodied agent are modeled as tokens to predict. Despite their promising performance, it remains unclear if embodied sequence modeling leads to the emergence of internal representations that represent the environmental state information. A model that lacks abstract state representations would be liable to make decisions based on surface statistics which fail to generalize. We take the BabyAI environment, a grid world in which language-conditioned navigation tasks are performed, and build a sequence modeling Transformer, which takes a language instruction, a sequence of actions, and environmental observations as its inputs. In order to investigate the emergence of abstract state representations, we design a "blindfolded" navigation task, where only the initial environmental layout, the language instruction, and the action sequence to complete the task are available for training. Our probing results show that intermediate environmental layouts can be reasonably reconstructed from the internal activations of a trained model, and that language instructions play a role in the reconstruction accuracy. Our results suggest that many key features of state representations can emerge via embodied sequence modeling, supporting an optimistic outlook for applications of sequence modeling objectives to more complex embodied decision-making domains.