LLM with structured prompts reconstructs ISRS GN formulas and derives a novel multi-span approximation validated with mean absolute error below 0.109 dB.
Artificial intelligence in drug development,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Mathematical Reasoning Enhanced LLM for Formula Derivation: A Case Study on Fiber NLI Modellin
LLM with structured prompts reconstructs ISRS GN formulas and derives a novel multi-span approximation validated with mean absolute error below 0.109 dB.