Ecosystem
Jev Decision Index fixes a scoring bug affecting 37 models
Developing 3 sources
A community-maintained Jev reproduction index corrected its same-skill scoring and added a public-versus-private comparison.
What happened lately
The community-maintained Jev Decision Index updated its scoring after discovering that three same-skill test sets had been omitted for a group of models. Its October 12 changelog describes the correction and a new chart comparing public scores with private same-skill scores. The index is a useful map of open attempts to reproduce typed decision models, but its rankings remain the maintainers’ evaluation under their own methodology.