SlowGuess/ABForge-Qwen3-8B-Task2

TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kTool Calling:SupportedPublished:Jun 11, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

SlowGuess/ABForge-Qwen3-8B-Task2 is an 8 billion parameter Qwen3-based model developed by SlowGuess, specifically fine-tuned for generating detailed ablation experiment design plans. This model excels at producing objectives, setups, variants, fixed protocols, and metrics for ablation studies based on research paper contexts and goals. It was post-trained using a supervised fine-tuning (SFT) and rubric-guided GRPO pipeline, making it highly specialized for scientific experiment design.

Loading preview...

ABForge-Qwen3-8B-Task2 Overview

This model, developed by SlowGuess, is an 8 billion parameter variant of the Qwen3 architecture, specifically engineered for Task 2: Ablation Plan Generation. It is part of the ABForge project, which focuses on a post-training pipeline for paper-grounded ablation design.

Key Capabilities

  • Ablation Experiment Design: Given a research paper's context and a specific goal, the model generates comprehensive ablation experiment design plans. This includes:
    • Defining the objective of the ablation.
    • Outlining the experimental setup.
    • Proposing different variants for comparison.
    • Specifying fixed protocols.
    • Identifying relevant metrics for evaluation.
  • Specialized Training: The model underwent a unique post-training pipeline, starting with supervised fine-tuning (SFT) from the base Qwen/Qwen3-8B model, followed by rubric-guided GRPO (Generative Reinforcement Learning with Policy Optimization). This process leverages training data derived from CC-licensed research papers, available in the SlowGuess/abforge-data dataset.

Good For

  • Automating Ablation Study Design: Researchers and developers can use this model to quickly generate structured plans for ablation experiments, saving time in the experimental design phase.
  • Ensuring Controlled Experiments: The model's output emphasizes detailed, controlled experiment designs, which is crucial for robust scientific validation.
  • Scientific Research Assistance: It serves as a valuable tool for anyone needing to systematically break down and analyze the contributions of different components within a research system or model.

Finetunes
3 models