Open Models Almanac
Methodspost-trainingreward modelling

specific method · filed under post-training

Length penalty

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.

sources
4
models
3
labs adopt it
3
strongest
used

Documented in

Evidence

4 quotes from 4 sources, one card per model, each opening with its strongest.

How sources treat it
used 4

DeepSeek-V3.2

DeepSeekadopted1 source

For reasoning and agent tasks, we employ rule-based outcome reward, length penalty, and language consistency reward.

usedtraining objectivein DeepSeek-V3.2DeepSeek
↑