AI RESEARCH

[R] TriAttention: Efficient KV Cache Compression for Long-Context Reasoning

r/MachineLearning

Submitted by /u/Benlus [link] [comments]