Popular repositories Loading

1. MemoryWeave MemoryWeave Public

   Compress AI conversations by 80-90% while preserving context. Never lose conversation history. Secure compression. ChatGPT & more.

2. Validation-Results Validation-Results Public

3. llm-stress-response-divergence llm-stress-response-divergence Public

   Three distinct behavioral strategies emerge from identical LLM stress conditions

   HTML

4. ssm-aware-lora-finetuning ssm-aware-lora-finetuning Public

   Systematic study of LoRA fine-tuning strategies for IBM Granite 4.0-H-Micro (Mamba-2 + Transformer hybrid). Demonstrates the impact of architecture-aware target selection and SSM core parameter co-…

   Python