r
reflex 4B
Small open decision model on Qwen3.5: state and typed questions in, calibrated probabilities out. An MIT-licensed Jev re-creation.
0 members list reflex 4B, unchanged over the last 12 weeks, data as of 24 September 2026
Usage on Stackness
0membersunchanged over the last 12 weeks
Members who list reflex 4B in their Stack, by week.
Nobody has this tool in their stack yet.
Posts about reflex 4B
- How do you benchmark a System One model? What JevBench scores, and what calibration says it cannot
JevBench scores Jev-class decision models on four axes: accuracy above chance, calibration, speed and cost per 1,000 decisions. Its v1.4.1 board puts Jev first at 63.3. A day later, a post argued that no benchmark can make these probabilities trustworthy on your data. Both are right, and this post covers what to measure yourself.