KaraKaraWitch/Golddiamondgold-Paperbliteration-L33-70b is a 70 billion parameter language model, a specialized "abliteration" of the GoldDiamondGold-L33-70b base model. It employs a constrained optimization methodology to significantly reduce refusal behaviors while preserving factual knowledge and reasoning capabilities. This model achieves a KL Divergence of 0.0055, indicating high fidelity to the original model's "Textbook" knowledge, making it suitable for applications requiring reduced refusals without sacrificing core intelligence.
No reviews yet. Be the first to review!