AI.info

Jaejun Shim

Explore Jaejun Shim on AI.info.

• When2Think: Learning Difficulty-Aware Length Control for Efficient Hybrid Reasoning Models