OpenTrain AI
No verified implementation yet

Awakening the Sleeping Agent: Lean-Specific Agentic Data Reactivates General Tool Use in Goedel Prover

Jui-Hui Chung, Hongzhou Lin, Lai Jiang, Shange Tang, Chi Jin

April 9, 2026arXiv: 2604.08388
0 repos~a few hours to reproduce
arXiv PDF

Abstract

Heavy supervised fine-tuning on a target domain can strongly suppress capabilities that were present in the base model. We study this phenomenon in formal mathematics using Goedel-Prover-V2, an open-source model heavily trained on 1.8 million formal-math examples. After domain specialization, the model almost completely loses its ability to produce valid tool calls, even when explicitly instructed to use tools, dropp...

Best Implementation

Maintained implementation evidence is not confirmed for this paper yet.

Use the Implementation Status and Reproduction Path sections below for the current action plan.

Reproduction Path

Follow this baseline workflow to decide if this paper is worth immediate prototyping.

  1. 1

    This is primarily a method paper. Reproduce it within a maintained framework baseline instead of chasing paper-specific repos.

  2. 2

    Start with framework-native implementations (e.g. PyTorch optimizer module, Optax, or Transformers training loops).

  3. 3

    Replicate the paper ablation settings first, then compare against modern baselines.

Time to first repro: a few hoursNo maintained paper-verified implementation is currently available

Additional Implementations

No additional verified repositories beyond the primary recommendation.

Hugging Face Artifacts

No trustworthy direct or curated related Hugging Face artifacts were found yet.

Research Context