Topics / Machine Learning (Statistics) / Oct 19 - 25, 2026
Weekly issue
Top 20 Machine Learning (Statistics) Papers: Oct 19 - 25, 2026
- 1WhatIf-Bench: Benchmarking Intervention-Aware Time Series Forecasting
We introduce WhatIf-Bench, a benchmark measuring how well forecasters use covariates to forecast a target's response to future events.
- 2
- 3On Reparameterizing The Score Function
We introduce History-Reparameterized (HR) policy gradients, which replace the frozen prefix in the score function with a differentiable reparameterized or relaxed history, allowing gradients of downstream log-probabilities to propagate backward through the generated trajectory.
- 4Density-Level Backward Error for Low-NFE Generative Flows
We develop a density-level backward-error framework that identifies the component of a solver's modified-equation defect that changes the evolving marginal, while excluding solenoidal trajectory distortion that is invisible in Wasserstein tangent geometry.
- 5Reliable AI-Assisted Decisions Via Selective Auditing
We consider online monitoring of an AI system that produces a sequence of decisions.
Get papers like these the day they come out, not a week later.
Follow this category- 6Learning to Be Monitored: Causal Bounds and Intervention-Aware Post-Training Recipes for Chain-of-Thought Monitorability
With this framework, we study whether LLMs can be trained to make their CoTs easier to monitor.
- 7When to Follow The Teacher: Advantage-Based On-Policy Distillation
We introduce Advantage-Based On-Policy Distillation (AOPD), which weights teacher actions by their counterfactual advantage under the student.
- 8Which Model for Which Role? Cost-Efficient Pareto Recommendation for LLM Agentic Workflows
We introduce Cost-Coupled Gittins (CC-Gittins), a multi-objective extension of single-objective cost-aware Gittins search for Pareto recommendation.
- 9Dimension-free Regimes for Unadjusted Langevin in Ridge-separable Models: A Subspace Robust Wasserstein Analysis
The Unadjusted Langevin Algorithm (ULA) is widely used for high-dimensional Bayesian inference.
- 10High-Value Training Data from Noisy, Diverse Sources
We develop a scalable approach to learning dataweights: given a small, high-quality target dataset, we use metagradients through the training process to efficiently optimize training-example weights directly for performance on the target domain.
- 11One Act, Many Masks: Effect-Marginal Policy Optimization for Long-Horizon Agents
Our analysis reveals that credit-text pairing directly shapes policy-update directions, even when the policy context and executable effect are held fixed. Motivated by this finding, we propose Effect-Marginal Policy Optimization (EMPO), which aligns policy optimization with execution-level equivalence through an effect-level surrogate that couples updates across alternative textual realizations.
- 12AI Mindmeld: A Latent Interconnect Protocol
We connect three protein and three small molecule specialists to Gemma 3 (4B and 12B parameters) and Kimi K2.6 (1T total, 32B active), freezing all pretrained models.
- 13R2-GP: Variance-Calibrated Priors for Additive Gaussian Processes
To also scale to large sample sizes, we extend R2-GP to Hilbert space approximation methods, a framework for kernels exploiting their spectral decompositions for fast and efficient inference.
- 14Modular Bayesian Inference for Partially Misspecified Mixtures with Uncertain Class Labels
While full Bayesian mixture modelling provides a principled approach to joint uncertainty quantification in this setting, often the model is misspecified or less trustworthy for populations that are not the target of the inference, which can distort posterior inference for the parameters of interest. We propose a modular Bayesian framework that safeguards against such biases.
- 15Curvature-Scaled Sharpness-Aware Minimization: Convergence Beyond Smoothness
We then establish a necessary condition for uniform distortion control among perturbations along the gradient whose nonnegative multipliers depend only on its norm. Motivated by this condition, we propose Curvature-Scaled SAM (CS-SAM), which scales both the perturbation and the descent step by suitable curvature upper bounds, and prove that it uniformly bounds the distortion.
- 16A Jump-Modified Wasserstein Gradient Flow for Maximum Mean Discrepancy with Energy Kernels
We study a jump-modified Wasserstein gradient flow of the squared maximum mean discrepancy generated by the energy kernels $K(z)=-|z|^q$, $0<q<2$, in every dimension $d\ge2$ as well as the one-dimensional range $1\le q<2$.
- 17A Plug-in Estimator of The Non-iterative Conditional Expectation G-formula Using Long Short-term Memory Models
In practical applications, parametric models are commonly used to estimate the joint distribution of confounders, treatments, and outcomes over time.
- 18Diffusion Language Models for Perplexity-Aware Black-Box Attacks
In this paper, we demonstrate that the same continuous latent-space framework can be coupled directly to a learned generative distribution over language, a continuous diffusion language model (dLLM), which removes the artificial random mapping while preserving the standard-Gaussian latent formulation and the downstream sampling or optimization procedure.
- 19Width and Depth Limits Commute During Training
Under the same $1/\sqrt{depth}$ residual-branch scaling (Depth-$\mu$P) and the $\mu$P width parameterization, we prove that, for any fixed number of gradient-descent steps with gradient clipping, the network output converges in $L^2$ along every joint width--depth growth path as both width and depth tend to infinity.
- 20Near Minimax Rate for Quantile Distributional Reinforcement Learning
In this paper, we study quantile distributional policy evaluation, where the return distribution under a given policy is approximated by the quantile Bellman fixed point $\bm\eta_m$.