Guang Dai

Deforming Videos to Masks: Flow Matching for Referring Video Segmentation

FZOO: Fast Zeroth-Order Optimizer for Fine‑Tuning Large Language Models towards Adam‑Scale Speed

Representation Alignment for Diffusion Transformers without External Components

Sharper Analysis of Single-Loop Methods for Bilevel Optimization

MSCR: Exploring the Vulnerability of LLMs’ Mathematical Reasoning Abilities Using Multi-Source Candidate Replacement

Towards Understanding The Calibration Benefits of Sharpness-Aware Minimization

Stochastic Gaussian Zeroth-Order Optimization: Improved Convergence Analysis under Skewed Hessian Spectra

ReBiSA: Data Reweighting with Bilevel Optimization for Safety Alignment

Numerical Sensitivity and Robustness: Exploring the Flaws of Mathematical Reasoning in Large Language Models

MonoLift: Learning 3D Manipulation Policies from Monocular RGB via Distillation

Manifold Constraint Reduces Exposure Bias in Accelerated Diffusion Sampling

Second-Order Fine-Tuning without Pain for LLMs: A Hessian Informed Zeroth-Order Optimizer

ProAdvPrompter: A Two-Stage Journey to Effective Adversarial Prompting for LLMs

A Computation and Communication Efficient Projection-free Algorithm for Decentralized Constrained Optimization

DynaMind: Reasoning over Abstract Video Dynamics for Embodied Decision-Making

The Advancement in Stochastic Zeroth-Order Optimization: Mechanism of Accelerated Convergence of Gaussian Direction on Objectives with Skewed Hessian Eigenvalues