影响力指数
论文质量、代表作、近期表现、广度与样本量置信度综合计算
48.17/100
前 9.5%
全站排名 #6,085
发表论文6 篇
平均评分
年均产出6.0 篇/年
Eric J Michaud
研究方向
deep learning · theory · AI safety · interpretability · reinforcement learning · mechanistic interpretability · emergence
18
Sparse Feature Circuits: Discovering and Editing Interpretable Causal Graphs in Language Models
ICLR 2025Oral
三作22
Not All Language Model Features Are One-Dimensionally Linear
ICLR 2025Poster
二作19
Efficient Dictionary Learning with Switch Sparse Autoencoders
ICLR 2025Poster
三作18
On the creation of narrow AI: hierarchy and nonlocality of neural network skills
NeurIPS 2025Poster
一作7
Interpretable Patterns in Random Initialization Unveil Final Representation
ICLR 2025Rejected
三作6
The Geometry of Concepts: Sparse Autoencoder Feature Structure
ICLR 2025Withdrawn
二作