Model techniques map
Techniquespost-trainingsupervised fine-tuning

specific method · filed under post-training

SFT bootstrapping with synthetic data

An initial supervised fine-tuning stage that bootstraps capabilities using synthetic data.

sources
2
models
2
labs adopt it
2
strongest
used

How sources treat it

One count per evidence span, weakest treatment to strongest.

used 2

Documented in

Evidence

2 spans quoted from the sources, strongest treatment first.

To bootstrap post-training, we ran an initial SFT on synthetic data generated by open-weights models including Kimi K2.5.

usedpost trainingin InklingThinking Machines Lab

an SFT stage that bootstraps capabilities partly with synthetic data

usedpost trainingin Laguna S 2.1Poolside

Filed alongside

Other methods under post-training :: supervised fine-tuning.