Ledger C10–C12 match the permute-pad-bmm path and Switch aux/z-loss. Section 8 adds overview and component TikZ; MoE capacity is C_moe so it does not collide with KDA chunk size.
Standalone tree split from LLMRL/projects/kda. Includes Triton dt_bias backward fix, train_k3 --preset 0.5b, SFT, Docker runtime, and tests.