The paper introduces a cost‑aware stopping rule for Freeze‑Thaw Bayesian Optimization called EI stopping, which uses a one‑step lookahead to decide whether to continue evaluating a candidate. EI stopping compares the expected improvement of the next increment to its cost and stops when no improvement exceeds the cost, guaranteeing that it never performs worse than stopping immediately. The authors provide a theoretical certificate that identifies when EI stopping is premature and show through experiments that EI stopping performs comparably to a two‑step lookahead correction in high‑cost scenarios, while the correction outperforms EI stopping at low costs.
arXiv:2606. 21641v2 Announce Type: replace-cross Abstract: Large language models (LLMs) have been proposed as hyperparameter-optimization (HPO) advisors that "warm-start" search from prior knowledge, proposing strong configurations in very few evaluations.
arXiv:2601. 07094v2 Announce Type: replace-cross Abstract: Bayesian optimization (BO) iteratively fits a Gaussian process (GP) surrogate to accumulated evaluations and selects new queries via an acquisition function.
The paper introduces BELT (BELief-based Termination), a method that decouples outcome prediction from stopping control in large language models (LLMs). BELT learns beliefs about current and future answers from a frozen LLM’s hidden states and applies a linear stopping policy, enabling a single probe to adapt to various token‑price trade‑offs without retraining. Across five out‑of‑distribution math benchmarks, BELT cuts token usage by up to 51% while maintaining accuracy, and its lightweight probe outperforms stronger baselines in both efficiency and parameter count, also transferring effectively to coding and agentic tasks.
arXiv:2605.22561v2 Announce Type: replace Abstract: Bayesian optimization (BO) is a widely used iterative black-box optimization method that utilizes Gaussian process (GP) surrogate models. In practi...
arXiv:2607. 04356v1 Announce Type: new Abstract: Bayesian Optimization (BO) generally begins with an initialization phase: a batch of $n_0$ uninformed evaluations.
arXiv:2606. 05606v1 Announce Type: new Abstract: LLM post-training often relies on reinforcement learning methods that sample multiple rollouts per prompt, yet most existing approaches use a fixed rollout budget for every prompt, despite large differences in the training signal different prompts provide.
arXiv:2610.14996v1 Announce Type: new Abstract: Choosing a few robust algorithms or schedules can require evaluating every candidate under many adverse scenarios, even though an agent only needs to i...
The paper introduces SELECT, an algorithmic framework for satisficing regret minimization in bandit problems, achieving constant expected satisficing regret when a satisficing arm exists. A variant, SELECT‑LITE, further ensures a light‑tailed satisficing regret distribution while maintaining constant expected regret in the realizable case and sub‑linear standard regret otherwise. Experiments on synthetic data and a real‑world dynamic pricing scenario demonstrate the practical effectiveness of both algorithms.
The paper evaluates how well typed decision models, which output probability distributions, actually provide reliable guarantees. By comparing open-weight Laya and Jev 1.13 across 190 evaluations, the study finds that point‑estimate thresholds frequently break promised accuracy, while a Beta lower‑bound rule and Learn‑then‑Test approach maintain guarantees more consistently. The work also explores calibration across languages, the impact of noise, data requirements for fine‑tuning, and the limitations of confidence in predicting outcomes or providing faithful explanations.
arXiv:2606. 30852v1 Announce Type: new Abstract: Reasoning models spend different amounts of useful computation across instances, but it remains unclear when a learned stopping rule improves over simple confidence or convergence thresholds.
arXiv:2609.39123v1 Announce Type: new Abstract: When optimizing an expensive black-box function sequentially, as in hyperparameter optimization, we may want to stop once the best evaluated value is c...
arXiv:2608. 14761v1 Announce Type: cross Abstract: At a finite public-chance cut, counterfactual regret minimization (CFR) must choose how many outcomes to evaluate before each regret update.