3 papers
stat.ML2024
On Stronger Computational Separations Between Multimodal and Unimodal Machine Learning
Ari Karchmer
Recently, multimodal machine learning has enjoyed huge empirical success (e.g. GPT-4). Motivated to develop theoretical justification for this empirical success, Lu (NeurIPS '23, A…
cs.CC2023
Distributional PAC-Learning from Nisan's Natural Proofs
Ari Karchmer
Carmosino et al. (2016) demonstrated that natural proofs of circuit lower bounds for imply efficient algorithms for learning -circuits, but only over \textit{the uniform dis…
cs.LG2023
Agnostic Membership Query Learning with Nontrivial Savings: New Results, Techniques
Ari Karchmer
(Abridged) Designing computationally efficient algorithms in the agnostic learning model (Haussler, 1992; Kearns et al., 1994) is notoriously difficult. In this work, we consider a…