Fan Zhang
7 papers · Latest:
RobotEQ: Transitioning from Passive Intelligence to Active Intelligence in Embodied AI
RobotEQ introduces the first benchmark for active intelligence, assessing if embodied AI can understand and adhere to social norms without explicit commands.
AffectGPT-RL: Revealing Roles of Reinforcement Learning in Open-Vocabulary Emotion Recognition
AffectGPT-RL uses reinforcement learning to optimize non-differentiable metrics in open-vocabulary multimodal emotion recognition, achieving SOTA results.
Collective quantum state at the atomic limit
Collective quantum states, like Luttinger liquids, persist and show distinct coupling behavior even at atomic dimensions.
RoboAlign-R1: Distilled Multimodal Reward Alignment for Robot Video World Models
RoboAlign-R1 improves robot video world models by using reward-aligned post-training and stabilized long-horizon inference, boosting task consistency and realism.
COCONUT: Toward practical time-evolving Sun-to-Earth magnetohydrodynamic modeling
COCONUT is a new, efficient, single time-evolving MHD model that simulates Sun-to-Earth conditions, improving accuracy and simplifying the modeling pipeline.
MER 2026: From Discriminative Emotion Recognition to Generative Emotion Understanding
MER2026 expands emotion understanding challenges from discriminative recognition to generative analysis, introducing new tasks and multimodal approaches.
UDM-GRPO: Stable and Efficient Group Relative Policy Optimization for Uniform Discrete Diffusion Models
UDM-GRPO integrates Uniform Discrete Diffusion Models with RL using novel insights for stable and efficient policy optimization, achieving SOTA results.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.