con-cord/Mod3-with-ref
VISIONConcurrent Unit Cost:1Model Size:4.3BQuant:BF16Context Size:32kPublished:Jul 17, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold
con-cord/Mod3-with-ref is a 4.3 billion parameter medical LLM-as-a-Judge model based on the Gemma-3-4B architecture. Developed by con-cord, it is specifically fine-tuned for evaluating generated medical answers against predefined clinical criteria. This version operates with expert reference answers, making it suitable for robust medical response assessment.
Loading preview...
Mod3-with-ref: Medical LLM-as-a-Judge
Mod3-with-ref is a specialized medical language model, built upon the Gemma-3-4B architecture and containing 4.3 billion parameters. It functions as an LLM-as-a-Judge, specifically designed for the critical task of evaluating medical responses.
Key Capabilities
- Medical Response Evaluation: Fine-tuned to assess the quality and accuracy of AI-generated medical answers.
- Clinical Criteria Adherence: Evaluates responses based on predefined clinical standards and guidelines.
- Reference-Based Assessment: This particular version (
with-ref) leverages expert reference answers to enhance the evaluation process, providing a robust mechanism for judging medical content. - Frameworks: Utilizes Transformers, PEFT/LoRA, and TRL for its underlying framework.
Good For
- Automated Medical Content Review: Ideal for systems requiring automated assessment of medical information generated by other LLMs.
- Quality Assurance in Healthcare AI: Can be integrated into pipelines to ensure the reliability and safety of AI-driven medical advice or information.
- Research in Medical LLM Evaluation: Provides a specialized tool for researchers studying the efficacy and accuracy of medical language models.