The paper introduces Online Hyperparameter Optimization (OHPO), framing it as an infinitely many‑armed bandit problem over mixed and conditional search spaces. It proposes the IMABO framework, which couples any bandit policy with any oracle for proposing new configurations, and presents IMOSS—a restart‑free anytime policy with provable regret bounds. Experiments show that IMABO, combined with practical oracles such as TPE, an incumbent‑mutation oracle, and a pretrained tabular foundation model, outperforms random search across a range of settings from classical ML models to LLM‑based agents.
By Louis Abraham, Tuan-Anh Nguyen, Nicolas Devatine
arXiv:2605. 14953v2 Announce Type: replace Abstract: We address the problem of conformal selection, where an agent must select a minimal subset of options to ensure that at least one ``success'' is identified with a pre-specified target probability $\phi$.
By Sreenivas Gollapudi, Kostas Kollias, Kamesh Munagala, Ali Sinop
TailSFT is a simple modification to supervised fine‑tuning that filters out already well‑modeled sequences, concentrating learning on the tail of the data distribution. On the OLMo‑3 7B model, this approach improves pass@16 performance on math and coding tasks by up to 17% absolute and yields up to 4% absolute gains in subsequent GRPO reinforcement‑learning runs, with only minimal computational overhead. The authors also provide a lightweight diagnostic to identify settings where TailSFT is most beneficial and argue for a stage‑aware development strategy that evaluates intermediate checkpoints by their support for later training.
By Sadhika Malladi, Samy Jelassi, Dylan Foster, Jordan T. Ash, Akshay Krishnamurthy
The paper introduces Online Surrogate Repair (OSR), a closed‑loop algorithm that decouples the frequency of high‑fidelity evaluations from the length of an agent’s search by selectively updating a surrogate model with sparse, high‑fidelity data. An acquisition rule determines which candidate designs receive expensive evaluations, and the resulting labels refine the surrogate for subsequent episodes. Experiments on synthetic environments and the MADE benchmark show that OSR can reduce regret more efficiently than fixed‑surrogate approaches, requiring fewer oracle queries than high‑fidelity feedback after every episode.
By Xiaotang Feng, Philip Torr, Bruno Andreis
arXiv:2605. 26418v2 Announce Type: replace-cross Abstract: A properly calibrated rule-based autoscaler can beat every one of six mainstream deep reinforcement learning (DRL) algorithms on cost across every workload we test - so when, if ever, does DRL actually help?
By Guilin Zhang, Chuanyi Sun, Kai Zhao, Shahryar Sarkani, John Fossaceca
The paper presents a convergence framework for deep $V$‑learning over a finite horizon $H$, deriving explicit bounds on policy loss by decomposing the Bellman update error into six residuals. It shows how $L^s$ concentrability controls expected $L^1$ loss, quantifies the impact of shared sampling across horizon levels, and provides optimal and near‑optimal sample allocations for statistical error rates. The work also establishes sharp action‑gap bounds under a margin condition, transfers optimal‑gap results to frozen‑iterate gaps, and offers consistency guarantees for generative‑reset approximate‑ERM procedures with exact action scores.
By Yury Kolomeytsev
arXiv:2602. 05459v2 Announce Type: replace Abstract: Offline goal-conditioned reinforcement learning (GCRL) is typically benchmarked by the best tuned success rate of each method.
By Jan Malte T\"opperwien, Aditya Mohan, Marius Lindauer
arXiv:2510. 14807v3 Announce Type: replace Abstract: We revisit exploration collapse in reinforcement learning with verifiable rewards (RLVR), from the perspective of the \emph{candidate distribution} for next-token prediction.
By Ruotian Peng, Yi Ren, Zhouliang Yu, Weiyang Liu, Yandong Wen
arXiv:2607. 13389v1 Announce Type: new Abstract: Reinforcement Learning (RL) post-training is increasingly used to adapt foundation models for reasoning, planning, and feedback-driven robot-learning pipelines, but constrained post-training resources are often summarized by a single total FLOP budget.
By Patrick Wilhelm, Odej Kao
arXiv:2605. 05481v2 Announce Type: replace Abstract: We revisit a classic "chicken-and-egg" problem in reinforcement learning: to safely improve a policy, the value function must be accurate on the state-visitation distribution of the updated policy.
By Dillon Sandhu, Ronald Parr
Pre-training followed by fine-tuning has become the dominant recipe for learning performant policies, and in value-based reinforcement learning (RL) this raises a natural question: given a pretrained policy, should the Q-function be pretrained on offline data too? Conventional wisdom suggests it should, but recent results show that online RL with a randomly-initialized Q-function can result in highly performant and reliable policies without needing to pretrain the Q-function.
arXiv:2608. 15088v1 Announce Type: cross Abstract: Human-in-the-loop (HIL) online reinforcement learning for real robots must absorb human interventions quickly while continuing to improve beyond the human prior.
By Zihang Wang, Yishan Wang