Jev glossary
Evaluation & TrainingReinforcement learning for calibrated decisions
RLCD
TypeSafe's post-training method: reinforcement learning that optimizes for calibrated decisions and probabilities instead of generated text.
- Category
- Evaluation & Training
- Also known as
- Reinforcement learning for calibrated decisions
- Reading time
- 1 min
- Related terms
- 5
- Directory entries
- 14
- Docs
- docs.typesafe.ai
- Added
- 2026-09-24
Definition
RLCD is the third post-training path the AI primer describes, alongside RLHF and RLVR. It optimizes a different output contract: the model does not generate text, it returns decisions and probabilities, and higher probability should correspond to a greater chance that the answer is correct.
Calibration is the measurable result: outcomes assigned 0.2 should occur about 20% of the time across many predictions. That property is what makes Jev's confidence usable by code.
Tagstrainingcalibration
Related terms
Definitions that connect to this one.
RLHFThe human-preference training method that turned pretrained models into chatbots, and the method TypeSafe deliberately did not use.Evaluation & TrainingRLVRTraining against automatically verifiable rewards, which produced strong reasoning models that are slower and more expensive.Evaluation & TrainingCalibrationThe property that outcomes given a probability of 0.2 occur about 20% of the time, measured across many predictions.Answers & ConfidenceTypeSafe AIThe company behind Jev and the System One API, including the official SDKs, cookbooks, console, and published workflow evals.Core ConceptsMode droppingPreference optimization narrowing a model's output distribution toward a favored style and away from other valid outputs.Evaluation & Training
From the Jev directory
Directory entries that go deeper on this term.
Why TypeSafe trains decision models with RLCD instead of RLHF: calibrated probabilities where 0.2 outcomes happen about 20% of the time, and the case for machine-to-machine automation.
Jev Guides & ArticlesDocs#official#docs#evaluation
Official
TypeSafe AI's product site for Jev, its first System One Model: typed decisions with calibrated confidence, performance and pricing claims, a FAQ, and links to the docs, console, workflow evals, and launch post.
Jev Guides & ArticlesDocs#official#docs#models
Official
A training recipe for JEV-style System 1 decision models on Qwen bases: LoRA plus a 24-slot decision head distilled from Jev v1.13's output distribution, serving calibrated per-option probabilities through a TypeSafe-compatible endpoint.
Jev Models & Reimplementations#community#github#training
Community
shibing624/train-jevPython
shibing624/train-jevPythonA no-code training platform for judgment models: define a judgment task, import decision samples, build a training set, train candidate probabilities, verify calibration and publish a service. The README says phase P3 is in implementation with no real GPU evidence yet.
Jev Tools & CLIs#community#github#python
Community
cers-ai/SystemOneStudioPython
cers-ai/SystemOneStudioPythonAn open Apache-2.0 toolkit to train, benchmark and serve System One decision models on your own hardware, with Kenning as the flagship model. Models answer typed questions with calibrated probabilities behind a POST /v1/systemone endpoint.
Jev Repos & SDKs#community#github#python
Community
systemonedev/systemone-builderPython
systemonedev/systemone-builderPythonA tuner that trains language models to answer closed-set decisions with a complete probability distribution in one forward pass, using a tokenized multi-class Brier loss so only the correct choice is needed to train.
Jev Repos & SDKs#community#catalog-import#madewithjev
Community
liushiliushi/jevtunerPython
liushiliushi/jevtunerPython8 more matching entries in the full directory.
From the Jev AI model community
Posts from builders shipping with Jev right now.