Lab 5: decoding changes outputs, not knowledge¶
The lab samples repeatedly from one fixed logit vector. Greedy decoding always chooses the maximum. Lower temperature sharpens the distribution; higher temperature flattens it; top-k removes all but the highest-k candidates before sampling.
Key distinction¶
The logits are fixed. Decoding changes which candidate is selected, not the model weights or the evidence in context. A more random answer is not new knowledge, and a deterministic answer is not guaranteed truth.
Experiments¶
- Add a tied maximum and inspect greedy tie behavior.
- Compare top-k with nucleus/top-p sampling.
- Estimate empirical entropy from the draw counts.
- Repeat with several seeds and confidence intervals.
- Construct a distribution where the correct token is ranked second; explain what each decoding rule can and cannot repair.