AI.info

Muchen Li

Explore Muchen Li on AI.info.

• MoME: Mixture-of-Memory Embeddings for Context-Aware Sparse Lookup
• Test-Time Steering for Lossless Text Compression via Weighted Product of Experts