10 papers · 1 filter
The Right Lesson at the Right Step: Deriving Control Updates for Self-Evolving Agents
Yunhe Su, ZiYi Dong, Tong Yu +4
Self-evolving agents improve future behavior by reusing past experience, typically as global prompts, memories, or reflections. Yet these mechanisms rarely control where experience…
EvoSCM: Scientific Belief Revision Through Causal Model Evolution and Experimentation
Qing Zhao, Haowei Li, Weijian Deng +3
Scientific discovery depends on the ability to form hypotheses, test them through experiments, and revise them when evidence disagrees. Existing LLM agents support this process by…
Ground4D: Consistency-Aware 4D Reconstruction from Monocular Video
Qing Zhao, Weijian Deng, Pengxu Wei +1
Learning a 4D scene representation from a single monocular video that supports dynamic novel-view synthesis while maintaining faithful geometry over time remains challenging. Dynam…
Weak-to-Strong Elicitation via Mismatched Wrong Drafts
Wei Deng
We consider whether off-policy experience from a smaller, weaker model can elicit capability in a stronger learner that on-policy RL fine-tuning (e.g., GRPO) does not reach. We fin…
Language Generation as Optimal Control: Closed-Loop Diffusion in Latent Control Space
ZiYi Dong, Yuliang Huang, Weijian Deng +3
This work reformulates language generation as a stochastic optimal control problem, providing a unified theoretical perspective to analyze autoregressive and diffusion models and e…
When Preference Labels Fall Short: Aligning Diffusion Models from Real Data
Weiyan Chen, Weijian Deng, Yao Xiao +5
Preference alignment aims to guide generative models by learning from comparisons between preferred and non-preferred samples. In practice, most existing approaches rely on prefere…
Where Detectors Fail: Probing Generative Space for Generalizable AI-Generated Image Detection
Zijie Cao, Weijie Tu, Yao Xiao +3
Detecting AI-generated images (AIGI) remains challenging because detectors often fail to generalize to unseen generators. Although existing methods are trained on large datasets, t…
Inference-Time Alignment of Diffusion Models via Trust-Region Iterative Twisted Sequential Monte Carlo
Weixin Wang, Yu Yang, Wei Deng +1
We study inference-time alignment for diffusion-based generative models, aiming to steer a base model toward high-reward outputs without updating its weights. Recent Sequential Mon…
Progressive Split Mamba: Effective State Space Modelling for Image Restoration
Mohammed Hassanin, Nour Moustafa, Weijian Deng +1
Image restoration requires simultaneously preserving fine-grained local structures and maintaining long-range spatial coherence. While convolutional networks struggle with limited…
Unveiling Perceptual Artifacts: A Fine-Grained Benchmark for Interpretable AI-Generated Image Detection
Yao Xiao, Weiyan Chen, Jiahao Chen +8
Current AI-Generated Image (AIGI) detection approaches predominantly rely on binary classification to distinguish real from synthetic images, often lacking interpretable or convinc…