Balancing helpfulness and safety (harmlessness) is a critical challenge in aligning large language models (LLMs). Current approaches often decouple these two objectives, training separate preference models for helpfulness and safety, while framing safety as a constraint within a constrained Markov Decision Process (CMDP) framework. This paper identifies a potential issue when using the widely adopted expected safety constraints for LLM safety alignment, termed "safety compensation", where the constraints are satisfied on expectation, but individual prompts may trade off safety, resulting in some responses being overly restrictive while others remain unsafe. To address this issue, we propose Rectified Policy Optimization (RePO), which replaces the expected safety constraint with critical safety constraints imposed on every prompt. At the core of RePO is a policy update mechanism driven by rectified policy gradients, which penalizes the strict safety violation of every prompt, thereby enhancing safety across nearly all prompts. Our experiments demonstrate that RePO outperforms strong baseline methods and significantly enhances LLM safety alignment.

Residual connection has been extensively studied and widely applied at the model architecture level. However, its potential in the more challenging data-centric approaches remains unexplored. In this work, we introduce the concept of Data Residual Matching for the first time, leveraging data-level skip connections to facilitate data generation and mitigate data information vanishing. This approach maintains a balance between newly acquired knowledge through pixel space optimization and existing core local information identification within raw data modalities, specifically for the dataset distillation task. Furthermore, by incorporating training-time refinements, our method significantly improves computational efficiency, achieving superior performance while reducing training time and peak GPU memory usage by 50%. Consequently, the proposed method Fast and Accurate Data Residual Matching for Dataset Distillation (FADRM) establishes a new stateof-the-art, demonstrating substantial improvements over existing methods across multiple dataset benchmarks in both efficiency and effectiveness. For instance, with ResNet-18 as the student model and a 0.8% compression ratio on ImageNet-1K, the method achieves 48.4% test accuracy in single-model dataset distillation and 50.9% in multi-model dataset distillation, surpassing RDED by +6.4% and outperforming

Semi-supervised learning (SSL) enhances model performance by leveraging abundant unlabeled data alongside limited labeled data. As vision foundation models (VFMs) become central to modern vision applications, this paper revisits SSL in the context of these powerful pre-trained models. We conduct a systematic study on tasks where frozen VFMs underperform and reveal several key insights when fine-tuning them. First, parameter-efficient fine-tuning (PEFT) using only labeled data often surpasses traditional SSL methods--even without access to unlabeled data. Second, pseudo-labels generated by PEFT models offer valuable supervisory signals for unlabeled data, and different PEFT techniques yield complementary pseudo-labels. These findings motivate a simple yet effective SSL baseline for the VFM era: ensemble pseudo-labeling across diverse PEFT methods and VFM backbones.

Despite the remarkable advances of Large Language Models (LLMs) across diverse cognitive tasks, the rapid enhancement of these capabilities also introduces emergent deception behaviors that may induce severe risks in high-stakes deployments. More critically, the characterization of deception across realistic real-world scenarios remains underexplored. To bridge this gap, we establish DeceptionBench, the first benchmark that systematically evaluates how deceptive tendencies manifest across different societal domains, what their intrinsic behavioral patterns are, and how extrinsic factors affect them. Specifically, on the static count, the benchmark encompasses 150 meticulously designed scenarios in five domains, i.e., Economy, Healthcare, Education, Social Interaction, and Entertainment, with over 1,000 samples, providing sufficient empirical foundations for deception analysis. On the intrinsic dimension, we explore whether models exhibit self-interested egoistic tendencies or sycophantic behaviors that prioritize user appeasement. On the extrinsic dimension, we investigate how contextual factors modulate deceptive outputs under neutral conditions, reward-based incentivization, and coercive pressures. Moreover, we incorporate sustained multi-turn interaction loops to construct a more realistic simulation of real-world feedback dynamics. Extensive experiments across LLMs and Large Reasoning Models (LRMs) reveal critical vulnerabilities, particularly amplified deception under reinforcement dynamics, demonstrating that current models lack robust resistance to manipulative contextual cues and the urgent need for advanced safeguards against various deception behaviors.

Forecasting real-world time series requires modeling both short-term fluctuations and long-term evolutions, as these signals typically exhibit multiscale temporal structures. A core challenge lies in reconciling such dynamics: high-frequency seasonality demands local precision, while low-frequency trends require global robustness. However, most existing methods adopt a unified loss function across all temporal components, overlooking their structural differences. This misalignment often causes overfitting to seasonal noise or underfitting of long-term trends, leading to suboptimal forecasting performance. To address this issue, we propose a Physics-guided Multiscale Loss Framework (PMLF) that decomposes time series into seasonal and trend components and assigns component-specific objectives grounded in the distinct energy responses of oscillatory and drift dynamics. Specifically, we assign a quadratic loss to seasonal components, reflecting the quadratic potential energy profile of molecular vibration, while a logarithmic loss is used for trend components to capture the sublinear energy profile of molecular drift under sustained external forces. Furthermore, we introduce a softmax-based strategy that adaptively balances the unequal energetic responses of these two physical processes. Experiments on different public benchmarks show that PMLF improves the performance of diverse baselines, demonstrating the effectiveness of physics-guided loss design in modeling structural heterogeneity in time series forecasting.

Discovering symbolic Partial Differential Equation (PDE) from data is one of the most promising directions of modern scientific discovery. Effectively constructing an expressive yet concise hypothesis space and accurately evaluating expression values, however, remain challenging due to the exponential explosion with the spatial dimension and the noise in the measurements. To address these challenges, we propose the ABL-PDE approach that employs the Abductive Learning (ABL) framework to discover symbolic PDEs. By introducing a First-Order Logic (FOL) knowledge base, ABL-PDE can represent various PDEs, significantly constraining the hypothesis space without sacrificing expressive power, while also facilitating the incorporation of problem-specific knowledge. The proposed consistency optimization process establishes a synergistic interaction between the knowledge base and the neural network learning module, achieving robust structure identification, accurate coefficient estimation, and enhanced stability against hyperparameter variation. Experimental results on three benchmarks across different noise levels demonstrate the effectiveness of our approach in PDE discovery.

Reinforcement Learning with Verifiable Rewards (RLVR) has proven effective for training large language models (LLMs) on complex reasoning tasks, such as mathematical problem solving. A prerequisite for the scalability of RLVR is a high-quality problem set with precise and verifiable answers.

Many decision-making processes involve evaluating and selecting items, including scientific peer review, job hiring, school admissions, and investment decisions. These domains feature error-prone evaluations and uncertainty about outcomes, which undermine deterministic selection rules. Consequently, randomized selection mechanisms are gaining traction. However, current randomized approaches are ad hoc and, as we prove, inappropriate for their purported objectives. We propose a principled framework for randomized decision-making based on interval estimates of item quality. We introduce MERIT (Maximin Efficient Randomized Interval Top-k), which maximizes the worst-case expected number of top candidates selected under uncertainty represented by overlapping intervals. MERIT provides optimal resource allocation under an interpretable robustness notion. We develop a polynomial-time, practically efficient algorithm and prove our approach satisfies desirable axiomatic properties not guaranteed by existing methods. Experiments on synthetic peer review data from grant funding and conferences demonstrate that MERIT matches existing algorithms' expected utility under fully probabilistic models while outperforming them under our worst-case formulation.

AI-driven discovery can greatly reduce design time and enhance new therapeutics' effectiveness. Models using simulators explore broad design spaces but risk violating implicit constraints due to a lack of experimental priors. For example, in a new analysis across diverse models on the GuacaMol benchmark using supervised classifiers, over 60% of molecules proposed had a high probability of being mutagenic. In this work, we introduce Medex, a dataset of priors for design problems extracted from literature describing compounds used in lab settings. It is constructed with LLM pipelines for discovering therapeutic entities in relevant paragraphs and summarizing information in concise fair-use facts. Medex consists of 32.3 million pairs of natural language facts, and appropriate entity representations (i.e.

Geometric graph neural networks (GNNs) that respect E(3) symmetries have achieved strong performance on small molecule modeling, but they face scalability and expressiveness challenges when applied to large biomolecules such as RNA and proteins. These systems require models that can simultaneously capture fine-grained atomic interactions, long-range dependencies across spatially distant components, and biologically relevant hierarchical structure--such as atoms forming residues, which in turn form higher-order domains. Existing geometric GNNs, which typically operate exclusively in either Euclidean or Spherical Harmonics space, are limited in their ability to capture both the fine-scale atomic details and the long-range, symmetry-aware dependencies required for modeling the multi-scale structure of large biomolecules. We introduce DualEquiNet, a Dual-Space Hierarchical Equivariant Network that constructs complementary representations in both Euclidean and Spherical Harmonics spaces to capture local geometry and global symmetry-aware features. DualEquiNet employs bidirectional cross-space message passing and a novel Cross-Space Interaction Pooling mechanism to hierarchically aggregate atomic features into biologically meaningful units, such as residues, enabling efficient and expressive multi-scale modeling for large biomolecular systems. DualEquiNet achieves state-of-the-art performance on multiple existing benchmarks for RNA property prediction and protein modeling, and outperforms prior methods on two newly introduced 3D structural benchmarks demonstrating its broad effectiveness across a range of large biomolecule modeling tasks.