Rlcd - GoKawiil

Explainer details RLCD, the reward-modeling technique underlying Jev

A technical write-up describes RLCD (Reinforcement Learning from Calibrated Decisions) as an evolution of reward modeling, moving from scalar reward scores to pairwise preference models like PPRM, and finally to a schema-conditioned, multiway probability distribution. It explains that Jev implements this RLCD objective by adding typed outputs and parallel inference, effectively turning the reward model itself into the primary decision-making system rather than a hidden scoring layer behind a generator.

di-zhang-llm.github.io · 2026-09-24

TypeSafe's Jev model disputed over 'calibrated probability' claims

A technical blog post examines TypeSafe's new Jev system, a 'System One Model' that converts unstructured input into typed decisions with attached probability scores instead of generating text. The author argues Jev is genuinely useful because it works without requiring training data, unlike a fine-tuned BERT model, but disputes TypeSafe's claim that its RLCD training method produces probabilities that are truly calibrated across any dataset.

alexmolas.com · 2026-09-23