arXiv:2610.08719v1 Announce Type: new Abstract: Reusing the scores that select a Best-of-$N$ winner can overstate its expected reward. We study evaluation from a fixed matrix of $K$ independent score...
arXiv:2608.29308v1 Announce Type: cross Abstract: In metric social choice, each voter ranks a set of $m$ candidates by her distance to them in an unknown metric space. The cost of a candidate is its...
arXiv:2606. 14690v1 Announce Type: new Abstract: We study a \emph{max-risk} objective for active learning in a multi-group mean estimation $d$-armed bandits: a learner adaptively allocates a budget of $T$ samples across $d$ groups to minimize the worst-case uncertainty index $\max_{k\in[d]}\sigma_k^2/n_k$, where $\sigma_k$ is the standard deviation of the distribution of arm $d$, and $n_k$ is the number of times arm $d$ is sampled.
arXiv:2609.39229v1 Announce Type: cross Abstract: Automatic evaluation of faithfulness increasingly relies on a large language model acting as a judge, yet the most reliable judges are proprietary fr...
arXiv:2604.19851v2 Announce Type: replace-cross Abstract: Computational social choice studies how conflicting individual preferences are aggregated into a collective decision, with applications to el...
arXiv:2610. 09651v1 Announce Type: new Abstract: DP-SGD protects training data by adding Gaussian noise to clipped gradients.
arXiv:2610.35277v1 Announce Type: new Abstract: Embedding-based verification often calibrates a false-accept threshold from all impostor pairs among $m$ identities, although these pairs share identit...
DP-SGD protects training data by adding Gaussian noise to clipped gradients. The amount of noise is usually chosen by running a numerical privacy accountant inside a search. We study DP-SGD with rando...
arXiv:2607. 02104v2 Announce Type: replace Abstract: Large language models (LLMs) are increasingly used as cheap, scalable judges that compare candidate outputs pairwise.
arXiv:2610.34375v1 Announce Type: cross Abstract: Best-of-N (BoN) inference returns the candidate with the highest proxy reward, so a larger $N$ tests the proxy's extreme tail more than its average r...
The paper introduces a derandomization framework for stochastic majority vote classifiers, converting PAC‑Bayesian guarantees into deterministic majority vote guarantees. By applying disintegrated PAC‑Bayesian theory to the space of vote weight vectors, the authors derive two families of high‑probability generalization bounds for both data‑independent and data‑dependent ensembles. These bounds naturally lead to a self‑bounding learning algorithm that optimizes deterministic majority vote performance.
The paper investigates how many human annotators are equivalent to a panel of 32 large‑language‑model (LLM) judges. By comparing the panel’s label distributions to empirical human labels on three ChaosNLI tasks, the authors find two distinct effective panel sizes: distribution‑error matching yields effective sizes of 2.304, 3.750, and 3.445, while spectral matching gives 4.242, 6.459, and 6.499, indicating a 1.72–1.89× gap. The study also explores how spectral diversity, participation ratio, and panel composition affect effective size, and demonstrates that carefully chosen panels can outperform baseline accuracy while improving effective size.