Training with Pseudo-Code for Instruction Following

ArXi:2505.18011v2 Announce Type: replace-cross Despite rapid advances in the capabilities of Large Language Models (LLMs), they continue to struggle with following relatively simple and unambiguous instructions, particularly when compositional structure is involved. Recent work suggests that models may follow instructions effectively when they are expressed in pseudo-code rather than natural language. However, writing pseudo-code programs can be tedious, and relying on few-shot nstrations or inference-time code prompting is often unnatural for non-expert users of LLMs.