Tuned 26B diffusion model beats Jev after adding 512-token decode-time reasoning budget

rickasaurus · x · 2026-09-21

Developer pomterree tested test-time scaling on DJev by adding a 512-token budget to the default-off think knob, letting the model reason before returning a structured read. A tuned 26B diffusion model then trivially beats Jev, which already topped the author's earlier benchmark of major Jev variants.

Original post →

More from Models

Models channel →