palm-2-l-instruct scores 0.909 on Winogrande few-shot. Where are the symbolic rule systems that get such results? https://github.com/LudwigStumpp/llm-leaderboa

palm-2-l-instruct scores 0.909 on Winogrande few-shot. Where are the symbolic rule systems that get such results? https://github.com/LudwigStumpp/llm-leaderboard/blob/main/README.md https://github.com/LudwigStumpp/llm-leaderboard/blob/main/RE...