aiqbench analysis
This research introduces SCARED, the first method to ensure safety during parameter-free adaptation in In-Context Reinforcement Learning by enforcing strict cost budgets within a constrained Markov decision process framework. The approach dynamically adjusts agent behavior based on user-defined safety thresholds, demonstrating robust performance across benchmarks where traditional ICRL methods lack real-world deployability due to uncontrolled risks. By addressing the critical gap of test-time behavioral safety without requiring model retraining or updates, this work significantly advances the practical application of adaptive AI agents in high-stakes environments.
Safe In-Context Reinforcement Learning
arXiv:2509.25582v4 Announce Type: replace Abstract: In-context reinforcement learning (ICRL) is an emerging RL paradigm where an agent, after pretraining, can adapt to out-of-distribution test tasks without any parameter updates, instead relying on an expanding context of interaction history. While ICRL has shown impressive generalization, safety during this adaptation process remains unexplored, limiting its applicability in real-world deployments where test-time behavior is expected to be safe. In this work, we propose SCARED: Safe Contextual Adaptive Reinforcement via Exact-penalty Dual, the first method that promotes safe adaptation of ICRL under the constrained Markov decision process framework. During the parameter-update-free adaptation process, our agent not only maximizes the reward but also keeps the accumulated cost within a user-specified safety budget. We also demonstrate that the agent actively reacts to the safety budget; with a higher safety budget, the agent behaves more aggressively, and with a lower safety budget the agent behaves more conservatively. Across challenging benchmarks, SCARED consistently enables safe and robust in-context adaptation, outperforming existing ICRL and safe meta-RL baselines.
Abstract:In-context reinforcement learning (ICRL) is an emerging RL paradigm where an agent, after pretraining, can adapt to out-of-distribution test tasks without any parameter updates, instead relying on an expanding context of interaction history. While ICRL has shown impressive generalization, safety during this adaptation process remains unexplored, limiting its applicability in real-world deployments where test-time behavior is expected to be safe. In this work, we propose SCARED: Safe Contextual Adaptive Reinforcement via Exact-penalty Dual, the first method that promotes safe adaptation of ICRL under the constrained Markov decision process framework. During the parameter-update-free adaptation process, our agent not only maximizes the reward but also keeps the accumulated cost within a user-specified safety budget. We also demonstrate that the agent actively reacts to the safety budget; with a higher safety budget, the agent behaves more aggressively, and with a lower safety budget the agent behaves more conservatively. Across challenging benchmarks, SCARED consistently enables safe and robust in-context adaptation, outperforming existing ICRL and safe meta-RL baselines.
Machine Learning (cs.LG)
arXiv:2509.25582 [cs.LG]
(or arXiv:2509.25582v4 [cs.LG] for this version)
Full text under the publisher's open licence · 174 words
Related coverage
The Alignment Problem: Computer Science Meets Analytic Philosophy
AI Alignment Forum · Oct 18, 2026
Continual learning might make your blocking monitors nearly useless
AI Alignment Forum · Sep 24, 2026
Adjustment Speed as a Safety Constraint for Nonstationary Reinforcement Learning
arXiv · Jul 27, 2026
AI alone won’t change your business. The system running it will.
Microsoft · Jun 2, 2026