The paper introduces a new approach to safety in contextual bandits with continuous actions, focusing on high‑probability constraints on the realized cost rather than expected cost. It presents the High‑Probability Constrained UCB algorithm, which balances reward exploration with conservative safety estimation, and provides theoretical regret guarantees for linear models and extensions to general function classes. Experiments demonstrate that this realized‑cost safety framework significantly reduces safety violations compared to expected‑cost constrained methods.
By Spyros Dragazis, Aldo Pacchiano
arXiv:2606. 15877v1 Announce Type: cross Abstract: Chain-of-thought (CoT) improves large language models' performance in math and symbolic reasoning.
By Alex Bogdan
The paper introduces a new approach to safety in contextual bandits with continuous actions by enforcing high‑probability constraints on the realized cost rather than on its expectation. It proposes the High‑Probability Constrained UCB algorithm, which balances optimistic reward exploration with pessimistic safety estimation. The authors provide theoretical regret guarantees for linear models and extend the analysis to general function classes, demonstrating experimentally that realized‑cost constraints significantly reduce safety violations compared to expected‑cost baselines.
The paper introduces a Bayesian self‑escalation strategy for hierarchical large‑language‑model agents, allowing an agent to detect during its own reasoning that it is unlikely to succeed and hand control over to a stronger model. The authors formalise this as an optimal‑stopping problem over a learned competence posterior, derive a myopic escalation threshold, and prove that the optimal policy is a time‑varying threshold without assumptions on the raw signal. They provide theoretical guarantees—including a 1/√n regret decay with n calibration trajectories—and validate the approach in simulations and a real‑model code‑generation cascade, showing that the escalation frontier outperforms post‑hoc routing at equal cost.
whyItMatters":"The study offers a principled, theoretically grounded method for agents to dynamically decide when to seek stronger models, potentially improving efficiency and reliability in hierarchical LLM systems."
By Nadeem Shaikh
arXiv:2607. 02206v1 Announce Type: cross Abstract: Predictions are increasingly used to guide high-stakes decisions, from treatment selection to policy making.
By Yurui Zheng, Ying Jin
The paper introduces Risk‑Sensitive Satisficing (RS), a bandit learning rule that evaluates actions based on their past frequency and their surplus or shortfall relative to an aspiration level. RS can be reformulated as Attention Competitive Bayes (ACB), a Bayesian‑like estimator where the aspiration level serves as a prior mean and the limited number of trials limits how strongly evidence can override this prior. ACB predicts that unchosen actions gradually revert toward the aspiration level (pseudo‑forgetting) and that when the aspiration is unattainable, exploration never fully settles, reducing regret in simulations.
By Moto Shinriki, Yu Kono, Tatsuji Takahashi
arXiv:2608. 13209v1 Announce Type: cross Abstract: Many operational decisions are sequences of interventions under a cumulative resource limit, such as a maintenance schedule within a crew-hour budget.
By Minkyoung Kim, Beakcheol Jang
NeuronSifter is a framework for planning interventions in central nervous system microenvironments by converting treatment regimens into state‑conditional target‑occupancy fields and propagating them through microenvironment dynamics. It selects measurements based on their expected reduction in intervention loss, integrating typed outcomes into a unified posterior. In synthetic Alzheimer’s disease scenarios, occupancy conditioning improves trajectory probability scores and intervention ordering accuracy, and decision‑directed acquisition reduces terminal risk compared to a Bayesian experimental design planner.
By Haowei Xu, Wanyi Fu, Hongbin Han, Zhaoheng Xie
The paper introduces MI‑SARSA, an on‑policy temporal‑difference algorithm that incorporates mutual‑information regularization to model bounded rationality in reinforcement learning. By penalizing state‑specific deviations from a learned marginal action prior, the algorithm selectively uses state information only when the expected return outweighs the informational cost, yielding a reward‑complexity tradeoff. MI‑SARSA also predicts reaction times, showing that stronger information penalties lead to simpler policies, lower control costs, and faster responses, while regularization mitigates performance loss after environmental shifts at the expense of asymptotic return.
By James Wu, Chris R. Sims
arXiv:2609.06667v1 Announce Type: new
Abstract: Adaptive dosing requires policies that reduce tumor burden without excessive toxicity. Learned dosing policies are typically judged against historical...
By Aleksandar Dimitrov, Giacomo Spigler
The paper introduces Cost-Aware Sequential Hypothesis Testing (CASHT), where a decision-maker selects sensing actions with varying random costs to identify the true hypothesis under an average-error constraint while minimizing expected total cost. For fixed costs, the optimal expected total cost scales as Θ(log(1/δ)) and can be achieved by Multihypothesis Sequential Probability Ratio Test-based procedures. The authors extend the framework to random costs under ex-post and ex-ante revelation models, analyze when action cancellation reduces cost, and demonstrate through simulations that CA variants consistently lower total cost compared to classical methods.
By George Vershinin, Asaf Cohen, Omer Gurewitz
arXiv:2610. 35877v1 Announce Type: cross Abstract: Changing a deployment policy can invalidate a world model's uncertainty while leaving the physical transition mechanism and predictor unchanged.
By Ayush Mittal, Dhruv Gupta