#llm-accuracy
#llm-accuracy

[ follow ]

#prompt-engineering #gpt-4o #prompt-politeness

Want better ChatGPT responses? Try this surprising trick, researchers say

Rude or abrasive prompt tones increased GPT-4o Deep Research mode accuracy compared with polite tones on moderate-to-high difficulty multiple-choice reasoning tasks.

[ Load more ]

#llm-accuracy#llm-accuracy

Want better ChatGPT responses? Try this surprising trick, researchers say

#llm-accuracy
#llm-accuracy