arXiv:2604.04089v3 Announce Type: replace-cross Abstract: Large language models can write scientific code, but direct paper-to-program translation remains fragile when correctness depends on tacit conventions in the literature. We identify this bottleneck as \textbf{knowledge externalization}: converting implicit computational assumptions -- index conventions, gauge choices, fermionic signs, contraction order, and memory constraints -- into an explicit technical specification before implementation. We evaluate a multi-stage, human-in-the-loop workflow that inserts such a specification, with va
Source: arXiv cs.AI — read the full report at the original publisher.