auto-paper-improvement-loop
wanshuiyin/Auto-claude-code-research-in-sleep
Improve a compiled academic paper through iterative LLM review, fix implementation, and recompilation across two rounds, using GPT-6-Astra via Codex with a fresh reviewer each round to avoid score inflation. Polishes writing quality, fixes theoretical inconsistencies, softens overclaims, and improves presentation after initial paper generation.
