Sorry "this beats GPT-4" - A new kind of LLM RANKINGS!!!

🏆 This leaderboard is based on the following three benchmarks.

Chatbot Arena - a crowdsourced, randomized battle platform. We use 100K+ user votes to compute Elo ratings.
MT-Bench - a set of challenging multi-turn questions. We use GPT-4 to grade the model responses.
MMLU (5-shot) - a test to measure a model's multitask accuracy on 57 tasks.

🔗 Links 🔗

ChatBOT Arena Leaderboard from Lmsys - https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard

Arena Leaderboard Elo Ranking Method - https://colab.research.google.com/drive/1RAWb22-PFNI-X1gPVzc927SGUdfr6nsR?usp=sharing

Play at the Arena - https://chat.lmsys.org/?arena …

arena benchmarks chatbot compute elo gpt gpt-4 kind leaderboard llm mmlu platform questions rankings responses set test