Showing 2026Show all
2 papers · 1 filter
cs.AI2026
Self-Play Pretraining with Zero Data
Aditya Cowsik, Kfir Dolev, Michael Y. Li +4
Advances in language modeling have been driven by scaling pretraining on ever more data. Yet, the training data is still largely curated on the model's behalf. A more general appro…
cond-mat.dis-nn2026
Hierarchical Prototype Emergence in Modern Hopfield Models
Aditya Cowsik, Adithya Sriram
Hierarchical correlations are a universal feature of any realistic model of data, and the question of how associative memory models may learn these correlations and generalize beyo…