jayzou3773/less-is-moe-gemma-4-26b-a4b-gpqa-main-64-intdim-e-50
Same weights, repacked (fewer layers) vs google/gemma-4-26B-A4B. jayzou3773/less-is-moe-gemma-4-26b-a4b-gpqa-main-64-intdim-e-50 matches part of google/gemma-4-26B-A4B (597 of its 1013 tensors; the rest of google/gemma-4-26B-A4B has no counterpart). On the shared part: Same weights in different files (re-saved, re-sharded or new metadata).
Tensors changed: Embeddings 0/1, Norms 0/271, other 0/120, MLP 0/90, Attention 0/115