This post is tracking the Argunauts project. We’re currently trying to teach LLMs logical argument analysis and argument mapping with Argdown, and are sharing progress and lessons learned in a series of articles.
1️⃣ Argunauts: Goals, Challenges, Masterplan 📘
2️⃣ Argunauts Training Phase I: Continual Pretraining on Synthetic Data 📘
3️⃣ Argunauts Training Phase II: Selfplay Finetuning Line-By-Line 📘