paper.dinq.meAI 与前沿科学顶会 · 学术新星雷达

更新于

返回新星榜

ZA

Zeynep Akata

ICML 2026×8可解释性多模态理论与学习动力学对齐与偏好效率与系统

新星分资深

33

一作贡献0

核心/参与22

方向广度10

合作网络12

0 一作 · 8 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

8

作者位次 #5/5ICML 2026Poster· Deep Learning / Generative Models and Autoencoders

The Latent Color Subspace: Emergent Order in High-Dimensional Chaos

Mateusz Pach · Jessica Bader · Quentin Bouniot · Serge Belongie · Zeynep Akata

理论与学习动力学OpenReview

作者位次 #5/5ICML 2026Spotlight· Deep Learning / Foundation Models

Dissecting Multimodal In-Context Learning: Modality Asymmetries and Circuit Dynamics in modern Transformers

Yiran Huang · Karsten Roth · Quentin Bouniot · Wenjia Xu · Zeynep Akata

可解释性多模态理论与学习动力学OpenReview

作者位次 #5/5ICML 2026Poster· General Machine Learning / Unsupervised and Semi-supervised Learning

SOTAlign: Semi-Supervised Alignment of Unimodal Vision and Language Models via Optimal Transport

Simon Roschmann · Paul KRZAKALA · Sonia Mazelet · Quentin Bouniot · Zeynep Akata

对齐与偏好多模态OpenReview

作者位次 #7/10ICML 2026Poster· Deep Learning / Generative Models and Autoencoders

Diamond Maps: Efficient Reward Alignment via Stochastic Flow Maps

Peter Holderrieth · Douglas Chen · Luca Eyring · Ishin Shah · Giri Anantharaman · Yutong He · Zeynep Akata · Tommi Jaakkola · Nicholas Boffi · Max Simchowitz

对齐与偏好效率与系统OpenReview

作者位次 #10/13ICML 2026Poster· Applications / Time Series

Mantis: Lightweight Foundation Model for Time Series Classification

Vasilii Feofanov · Songkang Wen · Shifeng Xie · Simon Roschmann · Marius Alonso · Hongbo Guo · Romain Ilbert · Malik TIOMOKO · Quentin Bouniot · Zeynep Akata · Lujia Pan · Jianfeng Zhang · Ievgen Redko

作者位次 #6/7ICML 2026Poster· Deep Learning / Foundation Models

Attentive Multi-Layer Fusion for Vision Transformers

Laure Ciernik · Marco Morik · Lukas Thede · Luca Eyring · Shinichi Nakajima · Zeynep Akata · Lukas Muttenthaler

多模态OpenReview

作者位次 #2/2ICML 2026Poster· Deep Learning / Generative Models and Autoencoders

Sparse Autoencoders are Topic Models

Leander Girrbach · Zeynep Akata

可解释性OpenReview

作者位次 #5/5ICML 2026Poster· Social Aspects / Accountability, Transparency, and Interpretability

TimeSAE: Causal Sparse Decoding for Faithful Explanations of Black-Box Time Series Models

Khalid Oublal · Quentin Bouniot · Qi Gan · Stephan Clemencon · Zeynep Akata

可解释性代码社会与人本OpenReview

常见合作者

QBQuentin Bouniot×5 SRSimon Roschmann×2 LELuca Eyring×2 MPMateusz Pach JBJessica Bader SBSerge Belongie YHYiran Huang KRKarsten Roth WXWenjia Xu PKPaul KRZAKALA SMSonia Mazelet PHPeter Holderrieth