Pinned Loading

1. FiCoCo FiCoCo Public

   [AAAI 26'] This is the official pytorch implementation for paper: Filter, Correlate, Compress: Training-Free Token Reduction for MLLM Acceleration

   Python 46 2

2. InnovatorLM/Innovator-VL InnovatorLM/Innovator-VL Public

   Fully Open-source Multimodal Language Models for Science Discovery

   Python 175 12

3. STaR-KV STaR-KV Public

   Official implementation of STaR-KV: training-free spatio-temporal KV cache compression for efficient GUI vision-language models.

   Python 22

4. Shadow-Weave/HMS Shadow-Weave/HMS Public

   Python 795 47

5. MAPGD MAPGD Public

   [TMLR 26']Official implementation of "MAPGD: Multi-Agent Prompt Gradient Descent for Collaborative Prompt Optimization"

   Python 16

6. OVC OVC Public

   Official implementation of "From Visual Affinity to Learning Credit in Multimodal Reinforcement Learning."

   Python 6