trends.barnas.net

phenomenon · first seen 29 Oct 2026

neural thickets

The hypothesis that pretrained models sit in dense neighborhoods of task-specific experts reachable by random weight perturbation.

Mentions per week

0219 May11 Aug10 NovPapersCodeDiscussion

PapersCodeDiscussion

Earliest sources

Latest sources

Mentioned together with

self-consistencyrandom weight perturbationmodel merging