arXiv:2605.21996v1 Announce Type: cross Abstract: Supervised fine-tuning (SFT) on long teacher trajectories is the dominant way to instill investigation and reasoning in open software-engineering (SWE) agents. Since every retained response becomes an imitation target, the student inherits the final outcome and intermediate flaws, including ungrounded leaps and redundant loops. High-quality training data must be effective(each […]
LVDrive: Latent Visual Representation Enhanced Vision-Language-Action Autonomous Driving Model
arXiv:2605.22089v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models have emerged as a promising framework for end-to-end autonomous driving. However, existing VLAs typically rely on sparse action supervision, which underutilizes their powerful scene understanding and reasoning capabilities. Recent attempts to incorporate dense visual supervision via world modeling often overemphasize pixel-level image reconstruction, neglecting semantically meaningful scene […]
Assessing global drivers of forest transpiration using clustered machine learning models
arXiv:2605.22755v1 Announce Type: new Abstract: Understanding the environmental drivers of forest transpiration is critical for improving global predictions of water availability and ecosystem health. Due to many competing controls on plant water stress and ecosystem transpiration, however, these drivers may vary widely across tree species which have adapted hydraulically to local climate conditions. Here, clustered […]
High-speed Networking for Giga-Scale AI Factories
arXiv:2605.21187v1 Announce Type: cross Abstract: As distributed model training scales to span hundreds of thousands of GPUs, scale-out networks face unprecedented performance and efficiency demands. NVIDIA Spectrum-X Ethernet has been designed from the ground up to achieve predictable and stable network performance with high utilization and low latency. This paper presents the Spectrum-X multiplane architecture, […]
Predicting Performance of Symbolic and Prompt Programs with Examples
arXiv:2605.21515v1 Announce Type: cross Abstract: LLM prompting is widely used for naturally stated tasks, yet it is unreliable it may succeed on a few test cases but fail at deployment time. We study performance prediction: given a program, either symbolic (e.g. Python) or a prompt executed on an LLM, and a few in-domain examples, predict […]
Scalable On-Policy Reinforcement Learning via Adaptive Batch Scaling
arXiv:2605.21557v1 Announce Type: cross Abstract: Conventional wisdom holds that large-batch training is fundamentally incompatible with Reinforcement Learning (RL) – beyond a modest threshold, increasing batch sizes typically yields diminishing returns or performance degradation due to the inherent non-stationarity of the data distribution. We challenge this view by observing that non-stationarity is not a fixed property […]
Amplifying, Not Learning: Fine-Tuned AI Text Detectors Amplify a Pretrained Direction
arXiv:2605.21653v1 Announce Type: cross Abstract: AI text detectors amplify a pretrained typicality axis; they do not construct an AI-vs-human boundary. On raw encoders before any task supervision, projecting onto centroid(AI)-centroid(HC3) achieves NYT-vs-HC3 AUROC 0.806/0.944/0.834 across three architectures (86-106% of the fine-tuned discrimination ceiling: on RoBERTa-base, raw projection exceeds fine-tuning); on RoBERTa-base, full fine-tuning reduces discrimination […]
Probabilistic Attribution For Large Language Models
arXiv:2605.21726v1 Announce Type: cross Abstract: The generative nature of Large Language Models (LLMs) is reflected in the conditional probabilities they compute to sample each response token given the previous tokens. These probabilities encode the distributional structure that the model learns in training and exploits in inference. In this work, we use these probabilities to situate […]
Comparing LLM and Fine-Tuned Model Performance on NVDRS Circumstance Extraction with Varying Prompt Complexity
arXiv:2605.21845v1 Announce Type: cross Abstract: Suicide is a leading cause of death in the United States, and understanding the circumstances that precede it requires extracting structured information from death investigation narratives. Many of these circumstances require semantic inference beyond simple keyword matching. We develop a “Complexity Score” algorithm that analyzes coding manual structure to predict […]
Engineering Hybrid Physics-Informed Neural Networks for Next-Generation Electricity Systems: A State-of-the-Art Review
arXiv:2605.21903v1 Announce Type: cross Abstract: The integration of machine learning with domain-specific physics is transforming the design, monitoring, and control of electricity systems, where data scarcity, limited interpretability, and the need to enforce physical laws constrain purely data-driven models. Physics-informed machine learning (PIML) addresses these limitations by embedding governing equations directly into the learning process, […]
Video as Natural Augmentation: Towards Unified AI-Generated Image and Video Detection
arXiv:2605.21977v1 Announce Type: cross Abstract: AI-generated content (AIGC) is rapidly improving, creating an urgent need for detectors that generalize across data sources, deployment pipelines, and visual modalities. A strongly generalizable detector should remain robust under distributional variations. However, we identify a consistent failure mode: SOTA AI-generated image detectors often collapse when applied to frames extracted […]
SDFStent: Real-time interactive virtual stenting via SDF deformation fields
arXiv:2605.22009v1 Announce Type: cross Abstract: Stenting is among the most common transcatheter interventions for congenital heart disease (CHD). Patient-specific computational fluid dynamics (CFD) simulations can predict hemodynamic outcomes of intervention scenarios but require post-operative vascular geometries that reflect stent-induced shape changes, which existing tools either model inadequately or require extensive time or manual effort to […]