mediumNewOpenAITogether AIFireworks AI

Implement a LoRA linear layer by hand, then decide whether to merge the adapter into a BF16 base and into an FP8 base.

The forward is three lines and still has a trap in the parenthesization. The merge is one line, and on a quantized base it can erase most of the fine-tune without an error message. The answer writes both, then measures what survives the rounding.

Updated Oct 2026 · Worked answers on inference mechanisms, with calculations, implementation details, and follow-up questions.

10 answers per topic instead of 5, plus saved progress and bookmarks · costs nothingor unlock all 213 remaining answers · $25

Every worked answer and derivation in all 12 topics, every concept page, the full practice-test bank, and every code drill’s edge-case tests and reference solution.

6 months · One payment · No auto-renewal

UP NEXT ON YOUR JOURNEY

Next in this trackImplement DeepSeek V3's MoE router from its config, then group 4,096 tokens by expert. What must the tests catch?Next in this trackImplement top-k sparse decode attention over a paged cache and prove it equals dense attention under the same mask.Next in this trackImplement one decode step of a Mamba-style layer, then make prefix reuse and speculative rollback work with checkpoints.

DISCUSSION · 0

Nobody has commented on this answer. Add a correction or another way to work it.