arXiv:2603.07514v2 Announce Type: replace-cross Abstract: Drifting models train one-step generators by optimizing a mean-shift discrepancy induced by a kernel between the data and model distributions, with Laplace kernels used by default in practice. At each point, this discrepancy compares the kernel-weighted displacement toward nearby data samples with the corresponding displacement toward nearby model samples, yielding […]
Can LLMs Reason Like Automated Theorem Provers for Rust Verification? VCoT-Bench: Evaluating via Verification Chain of Thought
arXiv:2603.18334v1 Announce Type: cross Abstract: As Large Language Models (LLMs) increasingly assist secure software development, their ability to meet the rigorous demands of Rust program verification remains unclear. Existing evaluations treat Rust verification as a black box, assessing models only by binary pass or fail outcomes for proof hints. This obscures whether models truly understand […]
Elastic Weight Consolidation Done Right for Continual Learning
arXiv:2603.18596v1 Announce Type: cross Abstract: Weight regularization methods in continual learning (CL) alleviate catastrophic forgetting by assessing and penalizing changes to important model weights. Elastic Weight Consolidation (EWC) is a foundational and widely used approach within this framework that estimates weight importance based on gradients. However, it has consistently shown suboptimal performance. In this paper, […]
TARo: Token-level Adaptive Routing for LLM Test-time Alignment
arXiv:2603.18411v1 Announce Type: cross Abstract: Large language models (LLMs) exhibit strong reasoning capabilities but typically require expensive post-training to reach high performance. Recent test-time alignment methods offer a lightweight alternative, but have been explored mainly for preference alignment rather than reasoning. To bridge this gap, we propose, Token-level Adaptive Routing (TARo), which steers frozen LLMs […]
Deep Expert Injection for Anchoring Retinal VLMs with Domain-Specific Knowledge
arXiv:2603.07131v3 Announce Type: replace-cross Abstract: Large Vision Language Models (LVLMs) show immense potential for automated ophthalmic diagnosis. However, their clinical deployment is severely hindered by lacking domain-specific knowledge. In this work, we identify two structural deficiencies hindering reliable medical reasoning: 1) the Perception Gap, where general-purpose visual encoders fail to resolve fine-grained pathological cues (e.g., […]
SODIUM: From Open Web Data to Queryable Databases
arXiv:2603.18447v1 Announce Type: cross Abstract: During research, domain experts often ask analytical questions whose answers require integrating data from a wide range of web sources. Thus, they must spend substantial effort searching, extracting, and organizing raw data before analysis can begin. We formalize this process as the SODIUM task, where we conceptualize open domains such […]
ICE: Intervention-Consistent Explanation Evaluation with Statistical Grounding for LLMs
arXiv:2603.18579v1 Announce Type: cross Abstract: Evaluating whether explanations faithfully reflect a model’s reasoning remains an open problem. Existing benchmarks use single interventions without statistical testing, making it impossible to distinguish genuine faithfulness from chance-level performance. We introduce ICE (Intervention-Consistent Explanation), a framework that compares explanations against matched random baselines via randomization tests under multiple intervention […]
CAFlow: Adaptive-Depth Single-Step Flow Matching for Efficient Histopathology Super-Resolution
arXiv:2603.18513v1 Announce Type: cross Abstract: In digital pathology, whole-slide images routinely exceed gigapixel resolution, making computationally intensive generative super-resolution (SR) impractical for routine deployment. We introduce CAFlow, an adaptive-depth single-step flow-matching framework that routes each image tile to the shallowest network exit that preserves reconstruction quality. CAFlow performs flow matching in pixel-unshuffled rearranged space, reducing […]
Farther the Shift, Sparser the Representation: Analyzing OOD Mechanisms in LLMs
arXiv:2603.03415v2 Announce Type: replace-cross Abstract: In this work, we investigate how Large Language Models (LLMs) adapt their internal representations when encountering inputs of increasing difficulty, quantified as the degree of out-of-distribution (OOD) shift. We reveal a consistent and quantifiable phenomenon: as task difficulty increases, whether through harder reasoning questions, longer contexts, or adding answer choices, […]
cuGenOpt: A GPU-Accelerated General-Purpose Metaheuristic Framework for Combinatorial Optimization
arXiv:2603.19163v1 Announce Type: new Abstract: Combinatorial optimization problems arise in logistics, scheduling, and resource allocation, yet existing approaches face a fundamental trade-off among generality, performance, and usability. We present cuGenOpt, a GPU-accelerated general-purpose metaheuristic framework that addresses all three dimensions simultaneously. At the engine level, cuGenOpt adopts a “one block evolves one solution” CUDA architecture […]
SpecForge: A Flexible and Efficient Open-Source Training Framework for Speculative Decoding
arXiv:2603.18567v1 Announce Type: cross Abstract: Large language models incur high inference latency due to sequential autoregressive decoding. Speculative decoding alleviates this bottleneck by using a lightweight draft model to propose multiple tokens for batched verification. However, its adoption has been limited by the lack of high-quality draft models and scalable training infrastructure. We introduce SpecForge, […]
Do Large Language Models Possess a Theory of Mind? A Comparative Evaluation Using the Strange Stories Paradigm
arXiv:2603.18007v1 Announce Type: cross Abstract: The study explores whether current Large Language Models (LLMs) exhibit Theory of Mind (ToM) capabilities — specifically, the ability to infer others’ beliefs, intentions, and emotions from text. Given that LLMs are trained on language data without social embodiment or access to other manifestations of mental representations, their apparent social-cognitive […]