LLM self-play on 20 Questions. gpt-3.5-turbo has a score of 68

🎉 ThursdAI Nov 30 // ChatGPT 1 year celebration special episode

Open-hermes 2.5 is better than GPT-3.5 by my real-world tests

I scored the top Open LLM Leaderboard models with my own benchmark

GPT-4 — LessWrong

ChatGPT 3.5 has officially reached, for me, worse than 13B quant

Conditional and Modal Reasoning in Large Language Models

Example Code & Implementation Considerations For GPT 3.5 Turbo

ChatGPT - Wikipedia

The correct rate of ChatGPT in the total exam and questions with

PDF) GPT-3.5/4 - Is the programming performance declining over time?

I scored the top Open LLM Leaderboard models with my own benchmark

Untitled, PDF, Graduate Record Examinations

PDF) GPT-3.5/4 - Is the programming performance declining over time?