Open Models Almanac
Methodspost-trainingsupervised fine-tuning

specific method · filed under post-training

Mask loss on incorrect tool calls and terminal actions

Excludes erroneous tool calls and terminal actions from the training loss.

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

sources
2
model
1
lab adopt it
1
strongest
used

Documented in

Aleph Alpha

Evidence

2 quotes from 2 sources, one card per model, each opening with its strongest.

How sources treat it
used 2

Kolibri-1

Aleph Alphaadopted2 sources

During training, we masked out erroneous tool calls and terminal actions, so the model learned to recover from errors without learning the erroneous actions themselves.

usedpost trainingin KolibriAleph Alpha
1 more from this model

mask the loss on wrong tool calls and terminal actions

usedunclearin KolibriAleph Alpha
↑