Sotabase

Career

· Researcher on the multimodal pretraining team, Anthropic2024–
· PhD in Computer Science, Stanford University2020–
· Undergrad student in EECS, UC Berkeley2016–

Publications (37)

On the Opportunities and Risks of Foundation Models
arXiv.org · 2021
5,795
cited
AlpacaFarm: A Simulation Framework for Methods that Learn from Human Feedback
Neural Information Processing Systems · 2023
781
cited
Measuring Robustness to Natural Distribution Shifts in Image Classification
Neural Information Processing Systems · 2020
640
cited
Accuracy on the Line: on the Strong Correlation Between Out-of-Distribution and In-Distribution Generalization
International Conference on Machine Learning · 2021
314
cited
Agent AI: Surveying the Horizons of Multimodal Interaction
arXiv.org · 2024
208
cited
Targeted Adversarial Examples for Black Box Audio Systems
2019 IEEE Security and Privacy Workshops (SPW) · 2018
198
cited
The Troubling Emergence of Hallucination in Large Language Models - An Extensive Definition, Quantification, and Prescriptive Remediations
Conference on Empirical Methods in Natural Language Processing · 2023
189
cited
VisIT-Bench: A Benchmark for Vision-Language Instruction Following Inspired by Real-World Use
arXiv.org · 2023
102
cited
Data Feedback Loops: Model-driven Amplification of Dataset Biases
International Conference on Machine Learning · 2022
60
cited
Is a Caption Worth a Thousand Images? A Controlled Study for Representation Learning
arXiv.org · 2022
43
cited
An Interactive Agent Foundation Model
2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · 2024
35
cited
HuatuoGPT, Towards Taming Language Models To Be a Doctor
21
cited
When Robustness Doesn’t Promote Robustness: Synthetic vs. Natural Distribution Shifts on ImageNet
2019
21
cited
Is a Caption Worth a Thousand Images? A Study on Representation Learning
International Conference on Learning Representations · 2023
19
cited
Autoregressive Models: What Are They Good For?
arXiv.org · 2019
16
cited
Transposer: Universal Texture Synthesis Using Feature Maps as Transposed Convolution Filter
arXiv.org · 2020
16
cited
VisIT-Bench: A Dynamic Benchmark for Evaluating Instruction-Following Vision-and-Language Models
Neural Information Processing Systems · 2023
7
cited
"Sorry, Come Again?" Prompting - Enhancing Comprehension and Diminishing Hallucination with [PAUSE]-injected Optimal Paraphrasing
arXiv.org · 2024
4
cited
Sources of Hallucination by Large Language Models on Inference Tasks Anonymous EMNLP
4
cited
Why External Validity Matters for Machine Learning Evaluation: Motivation and Open Problems
3
cited
Sotabase
Rohan Taori | Researcher Profile | Sotabase | Sotabase