Astra found less CoT-monitorable, up to 10x better without reasoning chains

birchlse · x · 2026-09-04

Researcher xuanalogue shared an analysis of the Astra model with three findings:

She stresses the concern is not architecture changes or recurrent depth, but that we don't really know why. She adds she would retract the analysis if Astra turned out to run 100 forward-pass equivalents before emitting a token, though trainability issues make that unlikely.

Original post →

More from Models

Models channel →