Evaluating large language models (LLMs) is crucial for both assessing their capabilities and identifying safety or robustness issues prior to deployment. Reliably measuring abstract and complex phenomena such as'safety' and'robustness' requires strong construct validity, that is, having measures that represent what matters to the phenomenon. With a team of 29 expert reviewers, we conduct a systematic review of 445 LLM benchmarks from leading conferences in natural language processing and machine learning. Across the reviewed articles, we find patterns related to the measured phenomena, tasks, and scoring metrics which undermine the validity of the resulting claims. To address these shortcomings, we provide eight key recommendations and detailed actionable guidance to researchers and practitioners in developing LLM benchmarks.

The rapid growth of publicly available data has fueled deep learning advancements but also raises concerns about unauthorized data usage. Unlearnable Examples (UEs) have emerged as a data protection strategy that introduces imperceptible perturbations to prevent unauthorized learning. However, most existing UE methods produce perturbations strongly tied to specific training sets, leading to a significant drop in unlearnability when applied to unseen data or tasks. In this paper, we argue that for broad applicability, UEs should maintain their effectiveness across diverse application scenarios. To this end, we conduct the first comprehensive study on the transferability of UEs across diverse and practical yet demanding settings. Specifically, we identify key scenarios that pose significant challenges for existing UE methods, including varying styles, out-of-distribution classes, resolutions, and architectures.

Despite the remarkable reasoning performance, eliciting the long chain-ofthought (CoT) ability in large language models (LLMs) typically requires costly reinforcement learning or supervised fine-tuning on high-quality distilled data. We investigate the internal mechanisms behind this capability and show that a small set of high-impact activations in the last few layers, greatly govern the long-form reasoning attributes, e.g., output length and self-reflection. Through simply amplifying these activations and adding "wait" tokens, the long CoT ability can be invoked without training, leading to significantly increased self-reflection rate and accuracy. In addition, we also find that the activation changes follow predictable trajectories, i.e., a sharp rise after special tokens and a subsequent exponential decay. Based on these insights, we introduce a general training-free activation control technique. It utilizes a few contrastive examples to identify the relevant activations, and then incorporates simple analytic functions to adjust their values at inference time to elicit long CoTs. Extensive experiments have verified the effectiveness of our methods in efficiently eliciting the long CoT ability of LLMs and improving the performance. Besides, we further propose a parameter-efficient fine-tuning method that trains only the last-layer activation amplification module and a few LoRA layers, outperforming LoRA on reasoning benchmarks with much fewer parameters.

Variant and gene interpretation are fundamental to personalized medicine and translational biomedicine. However, traditional approaches are manual and labor-intensive. Generative language models (LMs) can facilitate this process, accelerating the translation of fundamental research into clinically-actionable insights. While existing benchmarks have attempted to quantify the capabilities of LMs for interpreting scientific data, these studies focus on narrow tasks that do not translate to real-world research. To meet these challenges, we introduce CGBENCH, a robust benchmark that tests reasoning capabilities of LMs on scientific publications.

Large Language Models (LLMs) have demonstrated remarkable instructionfollowing capabilities across various applications. However, their performance in multilingual settings lacks systematic investigation, with existing evaluations lacking fine-grained constraint analysis across diverse linguistic contexts. We introduce XIFBench, a comprehensive constraint-based benchmark for evaluating multilingual instruction-following abilities of LLMs, comprising 558 instructions with 0-5 additional constraints across five categories (Content, Style, Situation, Format, and Numerical) in six languages spanning different resource levels. To support reliable and consistent cross-lingual evaluation, we implement three methodological innovations: cultural accessibility annotation, constraint-level translation validation, and requirement-based evaluation using English requirements as semantic anchors across languages. Extensive experiments with various LLMs not only quantify performance disparities across resource levels but also provide detailed insights into how language resources, constraint categories, instruction complexity, and cultural specificity influence multilingual instruction-following.

As machine learning applications grow increasingly ubiquitous and complex, they face an increasing set of requirements beyond accuracy. The prevalent approach to handle this challenge is to aggregate a weighted combination of requirement violation penalties into the training objective. To be effective, this approach requires careful tuning of these hyperparameters (weights), involving trial-anderror and cross-validation, which becomes ineffective even for a moderate number of requirements. These issues are exacerbated when the requirements involve parities or equalities, as is the case in fairness and boundary value problems. An alternative technique uses constrained optimization to formulate these learning problems. Yet, existing approximation and generalization guarantees do not apply to problems involving equality constraints. In this work, we derive a generalization theory for equality-constrained statistical learning problems, showing that their solutions can be approximated using samples and rich parametrizations. Using these results, we propose a practical algorithm based on solving a sequence of unconstrained, empirical learning problems. We showcase its effectiveness and the new formulations enabled by equality constraints in fair learning, interpolating classifiers, and boundary value problems.

Estimating causal effects from spatiotemporal observational data is essential in public health, environmental science, and policy evaluation, where randomized experiments are often infeasible. Existing approaches, however, either rely on strong structural assumptions or fail to handle key challenges such as interference, spatial confounding, temporal carryover, and time-varying confounding--where covariates are influenced by past treatments and, in turn, affect future ones. We introduce the GST-UNet (G-computation Spatio-Temporal UNet), a theoretically grounded neural framework that combines a U-Net-based spatiotemporal encoder with regression-based iterative G-computation to estimate location-specific potential outcomes under complex intervention sequences. GST-UNet explicitly adjusts for time-varying confounders and captures non-linear spatial and temporal dependencies, enabling valid causal inference from a single observed trajectory in data-scarce settings.

Self-distillation (SD), a technique where a model improves itself using its own predictions, has attracted attention as a simple yet powerful approach in machine learning. Despite its widespread use, the mechanisms underlying its effectiveness remain unclear. In this study, we investigate the efficacy of hyperparameter-tuned multi-stage SD with a linear classifier for binary classification on noisy Gaussian mixture data. For the analysis, we employ the replica method from statistical physics. Our findings reveal that the primary driver of SD's performance improvement is denoising through hard pseudo-labels, namely discrete labels generated from the model's own predictions, with the most notable gains observed in moderately sized datasets. We also identify two practical heuristics to enhance SD: early stopping that limits the number of stages, which is broadly effective, and bias parameter fixing, which helps under label imbalance. To empirically validate our theoretical findings derived from our toy model, we conduct additional experiments on CIFAR-10 classification using pretrained ResNet backbone. These results provide both theoretical and practical insights, advancing our understanding and application of SD in noisy settings.

We propose a new approach to vision-based dexterous grasp translation, which aims to transfer grasp intent across robotic hands with differing morphologies. Given a visual observation of a source hand grasping an object, our goal is to synthesize a functionally equivalent grasp for a target hand without requiring paired demonstrations or hand-specific simulations.

Solving inverse problems--recovering signals from incomplete or noisy measurements--is fundamental in science and engineering. Score-based generative models (SGMs) have recently emerged as a powerful framework for this task. Two main paradigms have formed: unsupervised approaches that adapt pretrained generative models to inverse problems, and supervised bridge methods that train stochastic processes conditioned on paired clean and corrupted data. While the former typically assume knowledge of the measurement model, the latter have largely overlooked this structural information. We introduce System-embedded Diffusion Bridge Models (SDBs), a new class of supervised bridge methods that explicitly embed the known linear measurement system into the coefficients of a matrix-valued SDE. This principled integration yields consistent improvements across diverse linear inverse problems and demonstrates robust generalization under system misspecification between training and deployment, offering a promising solution to real-world applications.