 Compute-Optimal LLMs Provably Generalize Better With ScaleInternational Conference on Learning Representations (ICLR), 2025 |
 Recursive PAC-Bayes: A Frequentist Approach to Sequential Prior Updates with No Information LossNeural Information Processing Systems (NeurIPS), 2024 |
 MMD-FUSE: Learning and Combining Kernels for Two-Sample Testing Without
Data SplittingNeural Information Processing Systems (NeurIPS), 2023 |
 Learning via Wasserstein-Based High Probability Generalisation BoundsNeural Information Processing Systems (NeurIPS), 2023 |
 The extended Ville's inequality for nonintegrable nonnegative
supermartingalesBernoulli (Bernoulli), 2023 |
 PAC-Bayes Bounds for Bandit Problems: A Survey and Experimental
ComparisonIEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2022 |
