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
56.78/100
前 5.8%
全站排名 #3,732
发表论文7 篇
平均评分
年均产出2.3 篇/年
Alessandro Stolfo
研究方向
Causality · causal mediation analysis · Reasoning · mathematical reasoning · arithmetics with LLMs · Interpretability · mechanistic interpretability · Math Word Problems · Transformer-based models · probing · learning dynamics
21
Transferring Linear Features Across Language Models With Model Stitching
NeurIPS 2025Spotlight
三作22
Dense SAE Latents Are Features, Not Bugs
NeurIPS 2025Poster
二作21
Improving Instruction-Following in Language Models through Activation Steering
ICLR 2025Poster
一作12
MIB: A Mechanistic Interpretability Benchmark
ICML 2025Poster