-
The Distribution Cliff: Why Hybrid Distillation Fails on Decoder-Only LLMs
-
Teaching AI to Embody Characters: A Replication of Open Character Training
-
Empirically Validating the Information-Theoretic Gap Between SL and RL
-
Constitutional AI from Base Models: Can You Train Safety Without Instruction Tuning?
-
GAN-Style Training for Jokes: When Metrics Lie
-
Seven Patterns From 300+ ML Evaluation Runs