arXiv:2607. 09765v1 Announce Type: new Abstract: A cheap swarm of unreliable agents can be steered to a correct consensus by a few strong, expensive "oracle" correctors.
arXiv:2607. 03436v1 Announce Type: new Abstract: Routing among large language models (LLMs) promises better quality at lower cost, motivated by the reported gap between learned routers and a per-instance oracle.
arXiv:2611.06372v1 Announce Type: new Abstract: When many inputs share a budget of repeated attempts by a large language model (LLM), a common rule gives harder inputs more. On public math, code and...
The paper investigates how verifier components used in post‑training and test‑time loops can collapse when they are refreshed based on aggregate success scores. It shows that as a policy improves on one conjunctive target, the verifier’s on‑policy AUC can shift to favor a worse component, leading to a preference for a full‑outcome predictor over the intended component oracle. The authors provide theoretical thresholds, empirical demonstrations on text control and ECG data, and propose a repair that audits the component on its own failures to prevent damage.
arXiv:2609.10196v1 Announce Type: cross Abstract: Attias, Hanneke and Ramaswami (NeurIPS 2025) asked whether randomization provably reduces the oracle calls needed for online learning when the class...
The study evaluates how to best split tasks among large‑language‑model agents for cross‑border VAT determination, comparing one broad agent to configurations ranging from one to five narrow agents. Across 4,400 runs—including token‑matched and failure‑injection scenarios—the intermediate configurations achieved the highest accuracy but did not surpass the fine‑endpoint benchmark, leaving the optimal decomposition hypothesis unconfirmed. The pilot provides a preregistered heuristic for right‑sizing decomposition, along with an oracle, dataset, and analysis pipeline.
arXiv:2610. 12211v1 Announce Type: new Abstract: A verifier that accepts or rejects whole answers reveals little: under a flat prior over $k$-bit answers, zero error needs $2^k-1$ verifications.
The paper shows that Worst‑Case Optimal Recovery (OR) and Bayesian learning solve the same Gaussian‑quadratic‑Hilbert problems, linking the radius of information to a nugget‑optimized Gaussian process posterior variance. It evaluates three Bayesian systems, demonstrating that OR can outperform Bayesian methods in certain calibration and reproducibility metrics, yet split‑conformal and other approaches can beat OR in interval scoring, especially under covariate shift. The authors propose matching the guarantee tool to the data regime and auditing that regime first.
arXiv:2502. 16744v3 Announce Type: replace Abstract: In adversarial Constrained Online Convex Optimization (COCO), a learner selects actions from a fixed convex set while seeking both low regret and low cumulative constraint violation (CCV) under time-varying constraints.
arXiv:2606. 02646v1 Announce Type: cross Abstract: Inference-time multi-agent LLM scaling lacks a shared unit: counting nominal agents conflates cost with independent evidence.
The paper proposes a claim‑specific verification audit for modular agents that replaces aggregate task scores with evidence‑based evaluations. Each agent conclusion is recorded with supporting evidence and classified as supported, unsupported, unresolved, or not evaluated, along with the boundary of validity. The audit employs three tools—oracle policies, perfect component replacements, and verifier‑score tests—to trace value changes, locate lost value, and assess verifier effectiveness, demonstrated on a portfolio‑allocation agent in a synthetic market.
arXiv:2608. 14795v1 Announce Type: new Abstract: An AI that can only give advice seems safe: the human is always free to ignore it.
arXiv:2608.12134v2 Announce Type: replace-cross Abstract: We study nonnegative submodular maximization on $n$ elements subject to a general matroid of rank $k$, when the offline algorithm is given an...