News
Duplicate 3 Layers in a 24B LLM and Logical Deduction Jumps from 0.22 to 0.76 — No Training Required
A researcher duplicated 3 specific layers in Devstral-24B and boosted logical deduction from 0.22 to 0.76 — no training, no weight changes. Here is how LLM Circuit Finder works and why it matters.