paper.dinq.meAI 与前沿科学顶会 · 学术新星雷达

更新于

返回新星榜

FL

Francesco Locatello

ICML 2026×6可解释性社会与人本对齐与偏好理论与学习动力学强化学习与后训练

新星分资深

24

一作贡献0

核心/参与12

方向广度10

合作网络6

0 一作 · 6 篇 · ICML 2026

获取完整研究档案

首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。

联系方式

实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。

2026 顶会论文

6

作者位次 #5/5ICML 2026Poster· General Machine Learning / Causality

Towards a Holistic Understanding of Selection Bias for Causal Effect Identification

Yiwen (Evie) Qiu · Filip Kovačević · Shimeng Huang · Peter Spirtes · Francesco Locatello

社会与人本OpenReview

作者位次 #5/5ICML 2026Poster· Applications / Chemistry, Physics, and Earth Sciences

The Perception–Physics Paradox: Probing Scientific Alignment with TC-Bench

Dingling Yao · Andrea Polesello · Adeel Pervez · Caroline Muller · Francesco Locatello

对齐与偏好可解释性OpenReview

作者位次 #3/3ICML 2026Poster· General Machine Learning / Causality

Addressing Instrument-Outcome Confounding in Mendelian Randomization through Representation Learning

Shimeng Huang · Matthew Robinson · Francesco Locatello

可解释性OpenReview

作者位次 #3/3ICML 2026Poster· Theory / Probabilistic Methods

Learning Discrete Diffusion on Graphs via Free-Energy Gradient Flows

Dario Rancati · Jan Maas · Francesco Locatello

理论与学习动力学OpenReview

作者位次 #3/3ICML 2026Poster· Deep Learning / Generative Models and Autoencoders

Toward Identifiable Sparse Autoencoders

Walter Nelson · Theofanis Karaletsos · Francesco Locatello

可解释性OpenReview

作者位次 #2/7ICML 2026Poster· Theory / Learning Theory

A Theory of How Pretraining Shapes Inductive Bias in Fine-Tuning

Nicolas Anguita · Francesco Locatello · Andrew Saxe · Marco Mondelli · Flavia Mancini · Samuel Lippl · Clémentine Dominé

强化学习与后训练数据与预训练社会与人本OpenReview

常见合作者

SHShimeng Huang×2 YQYiwen (Evie) Qiu FKFilip Kovačević PSPeter Spirtes DYDingling Yao APAndrea Polesello APAdeel Pervez CMCaroline Muller MRMatthew Robinson DRDario Rancati JMJan Maas WNWalter Nelson