← Lab
spacetax
confirmedPython · llama.cppModel behavior & reasoningJuly 7, 2026
the question
What does a single trailing space on a prompt do?
what came out
It changes a byte-level-BPE model's tokenization and greedy output on 100% of prompts, and flips a third of factual answers wrong -- on a real llama.cpp server. One trailing space. Pre-registered, independently verified.
method & receipts
- Result: confirmed
- Pre-registered — the prediction was written down and committed before the run.
- Reproducible — one script re-runs the whole thing from scratch.
- Independently reviewed — an adversarial review pass, kept in the repo.
- Machine-readable claims — every reported number, in a checkable file.
- Tested — a correctness/benchmark suite ships alongside the code.
→ read the code and re-run it
github.com/v-code01/spacetax