arXiv:2607. 19004v1 Announce Type: cross Abstract: We provide a simple and tight characterization of the types of inexact score oracle access that permit sampling with vanishing total variation bias, for a standard, well-behaved target family.
arXiv:2609.00691v1 Announce Type: new Abstract: Post-hoc out-of-distribution detectors are fitted on a finite reference set, so every score they produce is an estimate. If we had chosen a different s...
arXiv:2604.10857v2 Announce Type: replace-cross Abstract: Diffusion models generate samples by iteratively querying learned score estimates. A rapidly growing literature focuses on accelerating sampl...
The paper introduces Uniform Race (UR), a parameter‑free algorithm for approximate rejection sampling that selects a sample by maximizing the ratio of its importance weight to an independent uniform random variable. UR achieves the same total variation error bound as the optimal rejection sampler that requires knowledge of the best acceptance threshold, and it outperforms both the budget‑calibrated rejection sampler and sampling importance resampling in certain settings. The authors also provide theoretical conditions under which UR’s selection probabilities are uniquely optimal and validate the method on large‑language‑model reasoning tasks.
arXiv:2608. 06206v1 Announce Type: cross Abstract: Conformal prediction endows arbitrary black-box predictors with finite-sample, distribution-free marginal coverage, yet marginal validity can hide severe covariate-specific miscalibration, while exact distribution-free conditional coverage is finite-sample unattainable.
The study analyzes 373,019 judgments from LLM‑scored benchmarks, decomposing variance into system, item, judge, and interaction components via generalizability theory. It finds that with a single judge, generalizability converges to a ceiling determined by the system‑by‑judge variance, which is substantially lower in pairwise preference settings, allowing one judge to suffice. The research also reveals significant biases in presentation order and highlights that many published win‑rate claims fall below the measured floor of the benchmarks.
arXiv:2512. 24152v2 Announce Type: replace-cross Abstract: Sampling based on score diffusions has led to striking empirical results, and has attracted considerable attention from various research communities.
arXiv:2606. 11437v1 Announce Type: cross Abstract: Efficiently sampling from a complex probability distribution is a fundamental problem which has become increasingly pertinent in recent years with the rise of generative AI, as sophisticated sampling procedures from LLMs have been proposed to solve challenging reasoning problems.
arXiv:2608. 08424v1 Announce Type: cross Abstract: Conformal changepoint localization turns any score into a confidence set for the changepoint with finite-sample coverage.
arXiv:2604. 08438v2 Announce Type: replace Abstract: The Shapley value, and its broader family of semi-values, has received much attention in various attribution problems.
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...
arXiv:2606. 20820v2 Announce Type: replace Abstract: Can we trust evaluation scores to capture an LLM's true real-world performance?
The paper establishes the first polynomial lower bounds on the number of parallel rounds required for diffusion sampling with approximate scores. It proves a “~Ω(d^{1/3})” lower bound for sampling smooth, near‑isotropic Gaussian mixtures in ℝ^d, and an Ω(d) lower bound for uniform sampling from anisotropic axis‑aligned boxes inside the unit ball. These bounds apply to any randomized algorithm that can make polynomially many queries per round at arbitrary locations and noise levels, while maintaining inverse‑polynomial score error and constant total variation accuracy.