Open Models Almanac
Methodspost-trainingreward modelling

general family · filed under post-training

Reinforcement Learning with Verifiable Rewards

Trains policies using rewards determined by verifiable task outcomes rather than relying solely on subjective judgments.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See reward modelling for the methods filed beside it.

Also called Verifiable-reward reinforcement learning, verifiable reward, Verifiable rewards for RL environments.

sources
5
models
4
labs adopt it
3
strongest
core

Documented in

Evidence

5 quotes from 5 sources, one card per model, each opening with its strongest.

How sources treat it
optional 1used 3core 1

Kolibri-1

Aleph Alphaadopted1 source

All environments grade a sequence with a verifiable reward in [0, 1], except for the Hallucination Reduction environments, whose signed rewards lie in [−1, 1].

coretraining objectivein Kolibri RL training runAleph Alpha
↑