Pinned Loading

1. RetrieveReasoningNeurons RetrieveReasoningNeurons Public

   This work addresses identifying and enhancing reasoning-critical neurons in transformer MLP layers by computing neuron-level importance scores using multiple complementary methodologies.

   Python

2. Regularized-PEFT-of-Diffusion-Models-for-Artistic-Transfer Regularized-PEFT-of-Diffusion-Models-for-Artistic-Transfer Public

   This project aims to deliver a robust, parameter-efficient, and scalable fine-tuning method for high-fidelity style transfer in diffusion models. By integrating regularization through a frozen mode…

   Python

3. Demand-Paged-Virtual-Memory-Frame-Manager Demand-Paged-Virtual-Memory-Frame-Manager Public

   A C++ frame manager for operating system demand-paged virtual memory, using modular frame pools and efficient bitmap tracking to manage allocation and release of 32MB physical memory in x86 archite…

   C

4. MDLM_steering_with_policy MDLM_steering_with_policy Public

   Forked from anantmehta33/DLM_with_Policy

   Python

5. Triton-Flash-Attention-as-RL-Environment-for-Hardware-Aware-LLMs Triton-Flash-Attention-as-RL-Environment-for-Hardware-Aware-LLMs Public

   Designed a containerized RL environment to train LLMs to write hardware-optimized code, using continuous feedback based on real-time execution latency and strict mathematical correctness. Built a c…

   Jupyter Notebook

6. KV-cache-quantization KV-cache-quantization Public

   KV cache INT 8 quantization using AST graph compilation and kernel fusion.

   Jupyter Notebook