DeepSeek V4’s biggest omission

DeepSeek V4最大的遗憾

DeepSeek V4’s technical report omits Engram while listing mHC, CSA, HCA, Muon, and FP4. Engram was open-sourced by DeepSeek and Peking University in January, inserting lookup modules between Transformer layers 2 and 15; its 27B test raised MMLU by 3.4 and Multi-Query NIAH to 97.0%. The engineering signal is CXL pooling: 8 servers shared a 4TB memory pool with under 5% throughput loss.

Why it matters: HKR-H/K/R all pass: the omitted-Engram angle is clickable, with layer ranges, benchmark deltas, and CXL memory-pool numbers. It is analysis, not the V4 launch itself, so 78–84 fits.