Search

22 results for “MMLU” · page 2 of 2

1. Research · Efficiency & systems

   Tailored vector quantization space enables accurate 1-bit KV cache compression

   Tailoring the Quantization Space for 1-Bit KV Cache Compression

   TaSQ uses query-guided weighting, cross-head normalization, and covariance-aware grouping to improve VQ quality at extreme compression rates.

   MMLU: 71.26% accuracy (as reported)

   7 Oct 2026Seoul National University, Stanford University

2. Research · Efficiency & systems

   Two-rate precision allocation cuts MoE memory footprint and decode traffic

   DuoRate: Two-Rate Per-Expert Precision Allocation for Mixture-of-Experts Serving

   DuoRate stores weights once and adapts read precision per workload, matching 4-bit accuracy at 2.9 bpw

   MMLU: 81.47% accuracy (as reported)

   12 Oct 2026Huawei Technologies Ltd.