local LLMs, small LLMs, mixture of experts
Introducing Gemma 3 270M: The compact model for hyper-efficient AI
simonwillison.net·5d
Hierarchical Graph Feature Enhancement with Adaptive Frequency Modulation for Visual Recognition
arxiv.org·2d
ETTRL: Balancing Exploration and Exploitation in LLM Test-Time Reinforcement Learning Via Entropy Mechanism
arxiv.org·2d
Generalize across Homophily and Heterophily: Hybrid Spectral Graph Pre-Training and Prompt Tuning
arxiv.org·2d
Loading...Loading more...