AI.info

Masahiro Kaneko

Explore Masahiro Kaneko on AI.info.

• Autoregressive Direct Preference Optimization
• Stopping Computation for Converged Tokens in Masked Diffusion-LM Decoding