← Capabilities
Transformer architecture
not yet earnedI can implement attention and a transformer block from scratch and explain what each part is for without reciting a diagram.
What this claim costs
- 3 assessed items at tier 2 or above0 of 3 — 3 more to do, from 17 available
- A public artifact built for this claim — transformer-from-scratchNot built yet
- A documented negative result — where this stops workingNothing recorded. A capability with no known failure boundary has not been explored.
- Defended aloud, against questions you did not write1 available, none done
- Recalled cold, a week later, without notes5 scheduled, none done
The proof
transformer-from-scratch
A working implementation trained on something small, with the ablations that show which components matter.
What it has to contain
- Attention and a full decoder implemented without reference, running end to end
- Parameter count derived by hand and reconciled with the framework
- Ablations of at least three modern components, measured one at a time
- A reproduction attempt with its gap analysed and a residual left unexplained
Where it stops working. Not established yet. Anyone can ship a demo; the boundary is the part that is worth reading, so this claim is not earned until it is written down.
Skills underneath
- Attention
- Transformer architecture
- Implementing a transformer from scratch
17 assessed items across the specified courses evidence these, 0 done.