JAY ZENITH
I build post-training and evaluation systems for tool-using LLM agents: verifier-based environments, RL training infrastructure, and auxiliary learning objectives. I build PREDICT below.
- predict
Reactive vs. Predictive Post-Training for Coding Agents. I trained two Qwen3-4B coding agents on the same verifier-backed environment using Prime Intellect's Verifiers and PRIME-RL: reactive test-and-recover with GRPO versus predictive KEEP/REVISE with GRPO + auxiliary CE. Both improved with RL, but the predictive arm did not outperform. Trajectory analysis showed why: rollouts that chose REVISE almost never recovered, while cheap test feedback made direct observation more useful than foresight. [read the study] [code]