With a rubric-based optimized prompt, OpenAI o1 and gpt-4.1 scored 6/6 on all 13 NONMEM coding tasks, and every tested agent improved.
Acknowledgements This work is funded by a PharmAlliance Early Career Research Award and University of North Carolina at Chapel Hill
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
q-bio.OT 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
AI for NONMEM Coding in Pharmacometrics Research and Education: Shortcut or Pitfall?
With a rubric-based optimized prompt, OpenAI o1 and gpt-4.1 scored 6/6 on all 13 NONMEM coding tasks, and every tested agent improved.