Dani Yogatama

Research Scientist, Google DeepMind

Research Scientist, DeepMind

Joined

November 2017

Names

Dani Yogatama (Preferred)

Emails

****@gmail.com (Confirmed)

****@google.com (Confirmed)

****@baidu.com (Confirmed)

****@usc.edu (Confirmed)

Personal Links

DBLP

Career & Education History

Research Scientist

Google DeepMind (google.com)

2016 – Present

Research Scientist

DeepMind (google.com)

2016 – Present

Research Scientist

Baidu (baidu.com)

2015 – 2016

PhD student

Carnegie Mellon University (cmu.edu)

2010 – 2015

Advisors, Relations & Conflicts

No relations added

Expertise

No areas of expertise listed

Publications

The Semantic Hub Hypothesis: Language Models Share Semantic Representations Across Languages and Modalities
Zhaofeng Wu, Xinyan Velocity Yu, Dani Yogatama, Jiasen Lu, Yoon Kim
- ICLR 2025 Poster
- Readers: Everyone
DeLLMa: Decision Making Under Uncertainty with Large Language Models
Ollie Liu, Deqing Fu, Dani Yogatama, Willie Neiswanger
- ICLR 2025 Spotlight
- Readers: Everyone
Causal Interventions on Causal Paths: Mapping GPT-2's Reasoning From Syntax to Semantics
Isabelle Lee, Joshua Lum, Ziyi Liu, Dani Yogatama
- CaLM @NeurIPS 2024 Poster
- Readers: Everyone
(Pre-)training Dynamics: Scaling Generalization with First-Order Logic
Isabelle Lee, Daanish Shabbir, Nina Panickssery, Dani Yogatama
- ICLR 2025 Conference Withdrawn Submission
- Readers: Everyone
DeLLMa: A Framework for Decision Making Under Uncertainty with Large Language Models
Ollie Liu, Deqing Fu, Dani Yogatama, Willie Neiswanger
- OpenReview Archive Direct Upload
- Readers: Everyone
IsoBench: Benchmarking Multimodal Foundation Models on Isomorphic Representations
Deqing Fu, Ruohao Guo, Ghazal Khalighinejad, Ollie Liu, Bhuwan Dhingra, Dani Yogatama, Robin Jia, Willie Neiswanger
- COLM
- Readers: Everyone
Interpretable Diffusion via Information Decomposition
Xianghao Kong, Ollie Liu, Han Li, Dani Yogatama, Greg Ver Steeg
- ICLR 2024 poster
- Readers: Everyone
Understanding In-context Learning with a Pelican Soup Hypothesis
Ting-Rui Chiang, Dani Yogatama
- Submitted to ICLR 2024
- Readers: Everyone
Scaling Laws vs Model Architectures: How does Inductive Bias Influence Scaling?
Yi Tay, Mostafa Dehghani, Samira Abnar, Hyung Won Chung, William Fedus, Jinfeng Rao, Sharan Narang, Vinh Q. Tran, Dani Yogatama, Donald Metzler
- EMNLP 2023 Findings
- Readers: Everyone
The Distributional Hypothesis Does Not Fully Explain the Benefits of Masked Language Model Pretraining
Ting-Rui Chiang, Dani Yogatama
- EMNLP 2023 Main
- Readers: Everyone

View all 49 publications

Co-Authors

View all 171 co-authors