5 papers
Learning Correlated Reward Models: Statistical Barriers and Opportunities
Yeshwanth Cherapanamjeri, Constantinos Daskalakis, Gabriele Farina +1
Random Utility Models (RUMs) are a classical framework for modeling user preferences and play a key role in reward modeling for Reinforcement Learning from Human Feedback (RLHF). H…
The Space Complexity of Learning-Unlearning Algorithms
Yeshwanth Cherapanamjeri, Sumegha Garg, Nived Rajaraman +2
We study the memory complexity of machine unlearning algorithms that provide strong data deletion guarantees to the users. Formally, consider an algorithm for a particular learning…
Efficient Automated Circuit Discovery in Transformers using Contextual Decomposition
Aliyah R. Hsu, Georgia Zhou, Yeshwanth Cherapanamjeri +4
Automated mechanistic interpretation research has attracted great interest due to its potential to scale explanations of neural network internals to large models. Existing automate…
Heavy-tailed Contamination is Easier than Adversarial Contamination
Yeshwanth Cherapanamjeri, Daniel Lee
A large body of work in the statistics and computer science communities dating back to Huber (Huber, 1960) has led to statistically and computationally efficient outlier-robust est…
How much is a noisy image worth? Data Scaling Laws for Ambient Diffusion
Giannis Daras, Yeshwanth Cherapanamjeri, Constantinos Daskalakis
The quality of generative models depends on the quality of the data they are trained on. Creating large-scale, high-quality datasets is often expensive and sometimes impossible, e.…