Towards Monotonic Improvement in In-Context Reinforcement Learning

• URL: http://arxiv.org/abs/2509.23209v1
• Date: Sat, 27 Sep 2025 09:42:19 GMT
• Title: Towards Monotonic Improvement in In-Context Reinforcement Learning
• Authors: Wenhao Zhang, Shao Zhang, Xihuai Wang, Yang Li, Ying Wen,
• Abstract summary: In-Context Reinforcement Learning (ICRL) has emerged as a promising paradigm for developing agents that can rapidly adapt to new tasks.<n>Recent approaches train large sequence models on monotonic policy improvement data from online RL, aiming to a continue improved testing time performance.<n>We propose two methods for estimating Context Value at both training and testing time.
• Score: 18.67894044930047
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: In-Context Reinforcement Learning (ICRL) has emerged as a promising paradigm for developing agents that can rapidly adapt to new tasks by leveraging past experiences as context, without updating their parameters. Recent approaches train large sequence models on monotonic policy improvement data from online RL, aiming to a continue improved testing time performance. However, our experimental analysis reveals a critical flaw: these models cannot show a continue improvement like the training data during testing time. Theoretically, we identify this phenomenon as Contextual Ambiguity, where the model's own stochastic actions can generate an interaction history that misleadingly resembles that of a sub-optimal policy from the training data, initiating a vicious cycle of poor action selection. To resolve the Contextual Ambiguity, we introduce Context Value into training phase and propose Context Value Informed ICRL (CV-ICRL). CV-ICRL use Context Value as an explicit signal representing the ideal performance theoretically achievable by a policy given the current context. As the context expands, Context Value could include more task-relevant information, and therefore the ideal performance should be non-decreasing. We prove that the Context Value tightens the lower bound on the performance gap relative to an ideal, monotonically improving policy. We fruther propose two methods for estimating Context Value at both training and testing time. Experiments conducted on the Dark Room and Minigrid testbeds demonstrate that CV-ICRL effectively mitigates performance degradation and improves overall ICRL abilities across various tasks and environments. The source code and data of this paper are available at https://github.com/Bluixe/towards_monotonic_improvement .

Related papers

• Learning in Context, Guided by Choice: A Reward-Free Paradigm for Reinforcement Learning with Transformers [55.33468902405567]
  We propose a new learning paradigm, In-Context Preference-based Reinforcement Learning (ICPRL), in which both pretraining and deployment rely solely on preference feedback.<n>ICPRL enables strong in-context generalization to unseen tasks, achieving performance comparable to ICRL methods trained with full reward supervision.
  arXiv  Detail & Related papers  (2026-02-09T03:42:16Z)
• In-Context Reinforcement Learning From Suboptimal Historical Data [56.60512975858003]
  Transformer models have achieved remarkable empirical successes, largely due to their in-context learning capabilities.<n>We propose the Decision Importance Transformer framework, which emulates the actor-critic algorithm in an in-context manner.<n>Our results show that DIT achieves superior performance, particularly when the offline dataset contains suboptimal historical data.
  arXiv  Detail & Related papers  (2026-01-27T23:13:06Z)
• In-Context Reinforcement Learning through Bayesian Fusion of Context and Value Prior [53.21550098214227]
  In-context reinforcement learning promises fast adaptation to unseen environments without parameter updates.<n>We introduce SPICE, a Bayesian ICRL method that learns a prior over Q-values via deep ensemble and updates this prior at test-time.<n>We prove that SPICE achieves regret-optimal behaviour in both bandits and finite-horizon MDPs, even when pretrained only on suboptimal trajectories.
  arXiv  Detail & Related papers  (2026-01-06T13:41:31Z)
• Test-time Offline Reinforcement Learning on Goal-related Experience [50.94457794664909]
  Research in foundation models has shown that performance can be substantially improved through test-time training.<n>We propose a novel self-supervised data selection criterion, which selects transitions from an offline dataset according to their relevance to the current state.<n>Our goal-conditioned test-time training (GC-TTT) algorithm applies this routine in a receding-horizon fashion during evaluation, adapting the policy to the current trajectory as it is being rolled out.
  arXiv  Detail & Related papers  (2025-07-24T21:11:39Z)
• Scalable In-Context Q-Learning [68.9917436397079]
  We propose textbfScalable textbfIn-textbfContext textbfQ-textbfLearning (textbfSICQL) to steer in-context reinforcement learning.<n>textbfSICQL harnesses dynamic programming and world modeling to steer ICRL toward efficient reward and task generalization.
  arXiv  Detail & Related papers  (2025-06-02T04:21:56Z)
• A Snapshot of Influence: A Local Data Attribution Framework for Online Reinforcement Learning [37.62558445850573]
  We propose an algorithm, iterative influence-based filtering (IIF), for online RL training.<n>IIF reduces sample complexity, speeds up training, and achieves higher returns.<n>These results advance interpretability, efficiency, and effectiveness of online RL.
  arXiv  Detail & Related papers  (2025-05-25T19:25:57Z)
• Dynamic Loss-Based Sample Reweighting for Improved Large Language Model Pretraining [55.262510814326035]
  Existing reweighting strategies primarily focus on group-level data importance.<n>We introduce novel algorithms for dynamic, instance-level data reweighting.<n>Our framework allows us to devise reweighting strategies deprioritizing redundant or uninformative data.
  arXiv  Detail & Related papers  (2025-02-10T17:57:15Z)
• Emergence of In-Context Reinforcement Learning from Noise Distillation [46.29510499540939]
  We propose a new data acquisition approach that enables in-context Reinforcement Learning from noise-induced curriculum. We show that it is viable to construct a synthetic noise injection curriculum which helps to obtain learning histories. We experimentally demonstrate that it is possible to alleviate the need for generation using optimal policies, with in-context RL still able to outperform the best suboptimal policy in a learning dataset by a 2x margin.
  arXiv  Detail & Related papers  (2023-12-19T15:56:30Z)
• Statistically Efficient Variance Reduction with Double Policy Estimation for Off-Policy Evaluation in Sequence-Modeled Reinforcement Learning [53.97273491846883]
  We propose DPE: an RL algorithm that blends offline sequence modeling and offline reinforcement learning with Double Policy Estimation. We validate our method in multiple tasks of OpenAI Gym with D4RL benchmarks.
  arXiv  Detail & Related papers  (2023-08-28T20:46:07Z)
• On Context Distribution Shift in Task Representation Learning for Offline Meta RL [7.8317653074640186]
  We focus on context-based OMRL, specifically on the challenge of learning task representation for OMRL. To overcome this problem, we present a hard-sampling-based strategy to train a robust task context encoder.
  arXiv  Detail & Related papers  (2023-04-01T16:21:55Z)
• Curriculum Offline Imitation Learning [72.1015201041391]
  offline reinforcement learning tasks require the agent to learn from a pre-collected dataset with no further interactions with the environment. We propose textitCurriculum Offline Learning (COIL), which utilizes an experience picking strategy for imitating from adaptive neighboring policies with a higher return. On continuous control benchmarks, we compare COIL against both imitation-based and RL-based methods, showing that it not only avoids just learning a mediocre behavior on mixed datasets but is also even competitive with state-of-the-art offline RL methods.
  arXiv  Detail & Related papers  (2021-11-03T08:02:48Z)
• Off-Policy Meta-Reinforcement Learning Based on Feature Embedding Spaces [14.029933823101084]
  We propose a novel off-policy meta-RL method, embedding learning and evaluation of uncertainty (ELUE) ELUE learns a belief model over the embedding space and a belief-conditional policy and Q-function. We demonstrate that ELUE outperforms state-of-the-art meta RL methods through experiments on meta-RL benchmarks.
  arXiv  Detail & Related papers  (2021-01-06T05:51:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.