Song Mei
~Song_Mei1
14
论文总数
7.0
年均投稿
平均评分
接收情况11/14
会议分布
ICLR
6
NeurIPS
5
ICML
2
COLM
1
发表论文 (14 篇)
20259 篇
5
U-Nets as Belief Propagation: Efficient Classification, Denoising, and Diffusion in Generative Hierarchical Models
ICLR 2025Poster
4
A Statistical Theory of Contrastive Learning via Approximate Sufficient Statistics
NeurIPS 2025Poster
3
Choose Your Anchor Wisely: Effective Unlearning Diffusion Models via Concept Reconditioning
ICLR 2025Rejected
4
Implicit Bias of Gradient Descent for Non-Homogeneous Deep Networks
ICML 2025Poster
4
Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning
ICLR 2025Rejected
4
Improving LLM Safety Alignment with Dual-Objective Optimization
ICML 2025Poster
4
Active-Dormant Attention Heads: Mechanistically Demystifying Extreme-Token Phenomena in LLMs
ICLR 2025Rejected
4
Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning
NeurIPS 2025Poster
4
Generalization or Hallucination? Understanding Out-of-Context Reasoning in Transformers
NeurIPS 2025Poster
20245 篇
5
Statistical Estimation in the Spiked Tensor Model via the Quantum Approximate Optimization Algorithm
NeurIPS 2024Spotlight
4
How Do Transformers Learn In-Context Beyond Simple Functions? A Case Study on Learning with Representations
ICLR 2024Poster
4
Transformers as Decision Makers: Provable In-Context Reinforcement Learning via Supervised Pretraining
ICLR 2024Poster
4
Large Stepsize Gradient Descent for Non-Homogeneous Two-Layer Networks: Margin Improvement and Fast Optimization
NeurIPS 2024Poster
4
Negative Preference Optimization: From Catastrophic Collapse to Effective Unlearning
COLM 2024Poster