Dylan Slack
d[last name]@uci.edu

Hi! I'm a researcher at Google DeepMind focused on post-training research in GDM's GenAI team. I contributed to Gemini 2.5, Gemini 3.0, and Gemini 3.1. I also worked on GDM's AMIE medical AI system.

Previously, I recieved a Ph.D. from UC Irvine advised by Sameer Singh and Hima Lakkaraju. My thesis focused on making AI more reliable through natural language interaction, leveraging language models as a bridge between complex systems and human users.

CV /  Google Scholar /  Github /  X / 

profile photo
Selected Research

Full list of publications available on Google Scholar. * Denotes equal contribution.
Learning Goal-Conditioned Representations for Language Reward Models
Vaskar Nath*, Dylan Slack*, Jeff Da, Yuntao Ma, Hugh Zhang, Spencer Whitehead Sean Hendryx
NeurIPS, 2024
arXiv / code

A Careful Examination of Large Language Model Performance on Grade School Arithmetic
Hugh Zhang, Jeff Da, Dean Lee, Vaughn Robinson, Catherine Wu, Will Song, Tiffany Zhao, Pranav Raja, Dylan Slack, Qin Lyu, Sean Hendryx, Russell Kaplan, Summer Yue
NeurIPS D&B, 2024
arXiv

TalkToModel: Explaining Machine Learning Models with Interactive Natural Language Conversations
Dylan Slack, Satyapriya Krishna, Hima Lakkaraju*, Sameer Singh*
Nature Machine Intelligence, 2023
TSRML @ NeurIPS, 2022   Honorable Mention Outstanding Paper
Nature Link / arXiv / code / demo

Fooling LIME and SHAP: Adversarial Attacks on Post hoc Explanation Methods
Dylan Slack*, Sophie Hilgard*, Emily Jia, Sameer Singh, and Hima Lakkaraju
AIES, 2020   (Oral Presentation)
code / arXiv
Press: Deeplearning.ai / HBR

ยป See all publications on Google Scholar

Source modified from here.