cs.LGNov 5, 2026

How do Transformers Perform Low-Level Control In-Context?

Authors: Ebonye Smith, Aidan Andrews, Phoenix Wilson, Alexis Frias, Ayush Pandey, Gireeja Ranade

Organizations: Ebonye Smith (corresponding author), Phoenix Wilson, and Gireeja Ranade are with the Department of Electrical Engineering and Computer Sciences, University of California Berkeley, Berkeley, CA 94720, USA · Aidan Andrews is with the Physics Department, University of Illinois, Urbana-Champaign, Urbana, IL 61801 · Alexis Frias is with the Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA 90089 · Ayush Pandey is with the Department of Electrical Engineering, University of California Merced, Merced, CA 95343

Abstract

One of the remarkable abilities of Transformer-based neural networks is in-context learning. This motivates us to think about "in-context" control. Can we control a system knowing only (1) that it belongs to a particular family of systems and (2) some data on system evolution (provided as the context trajectory) but no specific parameter information? Specifically, we investigate the ability of Transformer-models to perform "in-context" control of unstable, nonlinear dynamical systems, where context consists only of the unforced system dynamics. We leverage a multi-head, GPT-2-style architecture, and we evaluate performance for two classical nonlinear systems -- the cartpole system and the two-link acrobot system, where the system parameters are unknown to the Transformer. We consider a setting where trajectories from an expert controller are available at training time but not at inference time. Here, we identify training techniques that are essential to train Transformers to perform low-level control. Through an analysis over checkpoints, we show that in-context control is possible, and the learned policy performance generally improves as context increases. Furthermore, the Transformer models can generalize to stabilize systems that are out-of-distribution.

Figures & tables

Explore similar work

CardsList
  1. Transformers for dynamical systems learn transfer operators in-context

    Feb 21, 2026Anthony Bao, Jeffrey Lai, William GilpinIn-Context LearningDynamical Systems

  2. Transformers Provably Implement In-Context Reinforcement Learning with Policy Improvement

    May 7, 2026Haodong Liang, Lifeng LaiReinforcement LearningTransformer-Based RL

  3. From Policy Evaluation to Off-Policy Control: Provable Emergence of In-Context Q-Learning

    Oct 13, 2026Yingjun Mo, Yajie Zhao, Fang KongIn-Context RLTransformer-Based RL