EpistemeAI/ReasoningCore-3B-RE1-V2C

TEXT GENERATIONConcurrency Cost:1Model Size:3.2BQuant:BF16Ctx Length:32kTool Calling:SupportedPublished:Feb 26, 2025License:llama3.2Architecture:Transformer0.0K Cold

ReasoningCore-3B-RE1-V2C is a 3 billion parameter, multilingual, reasoning-enhanced large language model developed by EpistemeAI. Built on an optimized transformer architecture, it is fine-tuned using Group Robust Preference Optimization (GRPO), supervised learning, and RLHF to excel at nuanced reasoning, dialogue management, retrieval, and summarization. This model supports 8 languages and features a 128k context length, making it suitable for complex conversational AI and knowledge-based applications.

Loading preview...

EpistemeAI/ReasoningCore-3B-RE1-V2C Overview

ReasoningCore-3B-RE1-V2C is a 3 billion parameter, multilingual, reasoning-enhanced large language model developed by EpistemeAI. It is built on an optimized transformer architecture and incorporates specialized reasoning pathways. The model has been extensively fine-tuned using Group Robust Preference Optimization (GRPO), supervised learning, and reinforcement learning with human feedback (RLHF) to align with human expectations for clarity, accuracy, and safety in complex tasks.

Key Capabilities & Features

  • Reasoning Enhancement: Specifically designed and fine-tuned with reasoning datasets (e.g., RUC-AIBOX/STILL-3-Preview-RL-Data, AI-MO/NuminaMath-TIR) to excel at nuanced reasoning.
  • Multilingual Support: Officially supports English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai, with pretraining on a broader range of languages.
  • Advanced Fine-tuning: Utilizes GRPO, a post-training reinforcement learning technique, to optimize for extended reasoning tasks, particularly mathematical problem-solving.
  • Context Length: Features a substantial 128k context window, enabling deep contextual understanding.
  • Optimized Architecture: Built on an optimized transformer architecture with specialized reasoning pathways.

Ideal Use Cases

  • Conversational AI: Excellent for assistant-like interactions requiring sophisticated dialogue management.
  • Knowledge Retrieval & Summarization: Capable of dynamic extraction and condensation of information.
  • Mobile AI-Powered Writing Assistants: Suitable for query reformulation and natural language generation in mobile environments.
  • General Natural Language Generation: Applicable to any task benefiting from advanced reasoning abilities, including complex problem-solving.