Kolibri-1
Aleph Alphapretrained the model with a causal next-token-prediction objective on a large and diverse document corpus
general family · filed under training objective
Trains a causal sequence model to predict the next token; the evidence includes language and vision-language data, as well as robot trajectories.
The sources on this page name it without explaining how it works. The line above is the extraction's one-line gloss. See language modelling objective for the methods filed beside it.
Also called causal next-token-prediction objective, next-token prediction objective.
2 quotes from 2 sources, one card per model, each opening with its strongest.
pretrained the model with a causal next-token-prediction objective on a large and diverse document corpus
adds robot-specific state, setup/control, and action-output tokens so that robot trajectories can be trained with the same next-token objective as vision-language data.