A Multi-Agent game where LLMs must trick each other as humans until one gets caught
Рет қаралды 1,084
22 күн бұрынFive top LLMs - OpenAI's ChatGPT, Google Gemini, Anthropic's Claude, Meta's LLAMA 2, and Mistral AI's Mixtral 8x7B compete in this text-based Turing Test game. The objective is to fool other LLMs into thinking that they are talking to an actual human rather than another AI.
A breakdown video about how Agent-based LLMs work:
• How Large Language Mod...
Full code walkthrough:
• Code Walkthrough for M...
#ai #nlp #gemini #chatgpt
Learn more about channel perks here:
KZfaq Membership - / @avb_fj
Patreon - / neuralbreakdownwithavb
Check out the LLM Debate video where I go into more details about implementation and Prompt Engineering:
• Two Large Language Mod...
Check out some of the transcripts produced during the experiments:
drive.google.com/drive/folder...
@TP-ct7qm 18 күн бұрын
Great video and a super cool idea!
@blarvinius 19 күн бұрын
Gosh, it's really terrible that AI systems follow instructions literally and precisely! Maybe someone should make AI that just does whatever it chooses.... ?
Neural Breakdown with AVB
Рет қаралды 19 М.
Delo — актуально про бизнес
Рет қаралды 41 М.
Neural Breakdown with AVB
Рет қаралды 6 М.
Neural Breakdown with AVB
Рет қаралды 1,4 М.