1
Why external scaffolds aren't enough. Chain-of-thought, self-consistency and search improve the outer procedure, but never check whether the model's internal pathway matches the logic each step needs.
A guided tour of LoTR
The problem, the idea, the method, and the results — an animated walkthrough of Logic-of-Thought Routing, built from the paper.
What the tutorial covers
2
How LoTR routes logic. A compact basis of recurring logic states, per-layer probes that read the current state, and cached head-templates that are softly blended into a gate matrix — online.
3
That it works and stays cheap. +7.93% average accuracy across 3 backbones × 8 benchmarks × 8 paradigms, with only +3.18% tokens and lower latency on Llama and Mixtral.