NeuralSirKrishna-7b: A Merged 7B Language Model

NeuralSirKrishna-7b is a 7 billion parameter language model developed by Kukedlc, created through a merge of three distinct models: NeuralKrishna-7B-v3, NeuralMarioMonarch-7B-slerp, and liminerity/M7-7b. This merge was performed using the dare_ties method, aiming to combine the strengths of its constituent models.

Key Capabilities

• Advanced Reasoning and Creative Generation: The model demonstrates a strong ability to generate complex and innovative theories, as showcased in its example response on the origin of life.
• Performance Benchmarks: On the Open LLM Leaderboard, NeuralSirKrishna-7b achieves an average score of 76.60. Notable individual scores include:
  • AI2 Reasoning Challenge (25-Shot): 73.72
  • HellaSwag (10-Shot): 89.05
  • MMLU (5-Shot): 64.63
  • TruthfulQA (0-shot): 75.60
  • Winogrande (5-shot): 85.32
  • GSM8k (5-shot): 71.27

Good For

• Complex Question Answering: Excels at generating detailed and imaginative responses to intricate prompts.
• Creative Content Generation: Suitable for tasks requiring novel ideas and disruptive concepts.
• Research and Development: Its strong reasoning capabilities make it a candidate for exploring advanced theoretical frameworks.