research note
This paper addresses the problem of human detection of synthetic speech (voice deepfakes) as a socio-technical phenomenon, emphasizing that human listeners must operate within environmental trust c…
research note
This paper investigates the adversarial robustness of forensic vision-language models (VLMs) that jointly detect image manipulations and generate natural language explanations justifying their judg…
research note
This paper addresses the challenge of efficiently specializing small computer-use agents (CUAs) to diverse software domains
research note
This paper addresses the longstanding problem of understanding the origin of quasi-periodic oscillations (QPOs) in accretion disks around black holes, focusing on the role of magnetic field configu…
research note
This paper addresses the challenge of aligning large language models (LLMs) toward desired behaviors—such as truthfulness, bias mitigation, and safety—by intervening on their internal representatio…
research note
PointQ-Bench addresses critical gaps in point cloud quality assessment (PCQA), moving beyond traditional scalar quality prediction to a comprehensive understanding that includes diagnostic defect i…
research note
This paper addresses key challenges in training Computer Use Agents (CUAs), which automate complex long-horizon GUI workflows across web and desktop environments
research note
This paper addresses the challenge of efficiently deploying Vision-Language-Action (VLA) models, which combine multi-billion-parameter language backbones and diffusion-based action heads
research note
This paper addresses the growing gap between text-to-image (T2I) generation capabilities and their evaluation metrics
research note
This paper tackles the problem of multimodal manipulation detection, specifically for forged image–text news pairs where existing methods fail to generalize well due to reliance on memorized artifa…
research note
This paper addresses key limitations in current sample-generation methods used for self-improving language models and agentic systems, specifically best-of-N sampling and tree search
research note
This paper addresses critical challenges in enabling effective visual search within Large Vision-Language Models (LVLMs), which are integral to the multimodal thinking-with-images paradigm
