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.
The paper investigates how self‑improving large language models (LLMs) can suffer from the winner’s curse when selecting better candidates based on noisy evaluation sets. By modeling correlated errors and empirically testing Qwen models that rewrite their own instructions, the authors find that most proposals after the first are harmful and that the reported gains on the selection set often overstate true held‑out performance. The study shows that larger selection sets can improve held‑out gains on some tasks (e.g., TREC) but not others (e.g., GSM8K), and that common acceptance rules do not outperform simple greedy selection across entire runs.
arXiv:2606. 31630v1 Announce Type: new Abstract: Language models increasingly write probabilistic programs (in NumPyro, Stan, or Pyro), but a program that compiles, runs, and passes every unit test can still be \emph{statistically} wrong -- a Gaussian likelihood for heavy-tailed data, a Poisson for over-dispersed counts, an invalid prior support, or a pathological parameterization.
StepCOPS is a new method for selecting a language‑model policy from many checkpoints, prompts, and decoding rules by providing closed‑testing lower‑tail certificates. It uses an independent proposal split to nominate a lower‑tail floor for each candidate, applies exact binomial tests on a fresh certification split, and employs Holm’s step‑down procedure to certify a set of floors. In experiments across 24 configurations and 11 benchmarks, StepCOPS achieves 96.4% selected‑policy coverage, raises the certified floor by 1.5 points over prior methods, stays 0.6 points below a large‑reference jury oracle, and abstains in 2.4% of trials.
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.
The paper introduces Evidence-Carrying Termination (ECT), a method that allows tool‑using large language model agents to declare a task complete only when a typed certificate links every required answer claim to valid, in‑scope trace evidence and a deterministic replay confirms the claimed value. In controlled experiments across 48 synthetic tasks and 576 trajectories, ECT eliminated unsafe completions and premature unsupported terminations, outperforming existing termination critics and controllers while maintaining comparable supported completion rates.
StrictPRMBench is a replayable evaluation benchmark designed to audit fixed‑score PRM aggregation rules. It standardizes mathematical reasoning pools, scorer evidence, and deletion interventions so that different reducers confront identical selection problems. Across seven checkpoints and two benchmarks, the benchmark reveals that product aggregation yields higher deletion‑induced selection instability than geometric mean, with statistically significant differences in most cells.
arXiv:2608. 07913v1 Announce Type: cross Abstract: Selective-risk certificates promise that accepted outputs meet a declared error target.
arXiv:2607. 02255v1 Announce Type: new Abstract: Memory for a long-horizon LLM agent is a contract about what each future decision is allowed to see.
arXiv:2607. 06503v1 Announce Type: new Abstract: Large language model (LLM) agents solving multi-step tasks frequently commit to trajectories that are doomed to fail, yet continue to consume substantial inference compute before the failure becomes observable.
arXiv:2609.13714v1 Announce Type: new Abstract: An updated model can improve an aggregate metric while degrading a slice that matters to a downstream user. We study checkpoint selection subject to no...
arXiv:2605. 22148v3 Announce Type: replace Abstract: A large language model (LLM) agent that writes and edits its own skill library must also decide which skills to keep, from one noisy scalar per skill.