LFM3-1.7B
Liquid AIA length penalty in the reward keeps answers concise.
1 more from this model
A length penalty in the reward keeps answers concise.
specific method · filed under post-training
Penalizes response length to encourage concise outputs; the evidence does not establish a single shared penalty formula.
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 length penalty reward, Reward length penalty for concise answers, Reward length penalty for concise outputs, Length-adjusted RL reward.
4 quotes from 4 sources, one card per model, each opening with its strongest.
A length penalty in the reward keeps answers concise.
A length penalty in the reward keeps answers concise.
length-based adjustments are applied to the RL rewards for them.