Tattoo Tool Grounded Thinking Prm

Build process reward models for tabular reasoning by grounding intermediate reasoning steps in executable tool operations. Train on 60K+ step-level annotations combining verification rationales with tool execution feedback. TaTToo achieves 30.9% improvement over policies using only final rewards, outperforming 72B reasoning models with just 8B parameters via better step-level supervision.

adu2021 Updated 7 repo stars

File contents

adu2021/skillxiv commit 1f97035b3e

Frequently asked questions

npx get-skill add adu2021/tattoo-tool-grounded-thinking-prm