trends.barnas.net

method · first seen 13 Oct 2026

chain-of-thought compression

Techniques that shorten long reasoning traces in reasoning models to reduce token usage without hurting accuracy.

Mentions per week

0219 May11 Aug10 NovPapersCodeDiscussion

PapersCodeDiscussion

Earliest sources

Latest sources

Mentioned together with

contrastive decodingreinforcement learning with advantage rescalingcausal intervention