Structured versus prose rendering of identical theory text did not change LLM-built programs as predicted; a 16-block randomized trial with exact randomization inference found the preregistered hypothesis unsupported.
The target is then fetched exactly once withGET, and its raw bytes are persisted before parsing
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Does the way we write a theory change the program an LLM builds from it? A prospective randomized study of renderer format in LLM theory-to-program translation
Structured versus prose rendering of identical theory text did not change LLM-built programs as predicted; a 16-block randomized trial with exact randomization inference found the preregistered hypothesis unsupported.