Skip to content

DEV Community

Broken Campus Series' Articles

Back to Torkian's Series

Cover image for The best score on my AI-agent benchmark is a refusal

Torkian

Aug 11

The best score on my AI-agent benchmark is a refusal

#ai #llm #agents #machinelearning

16 min read

Cover image for I ran 8 models through the same broken agent. If you're picking one, none win.

Torkian

Aug 11

I ran 8 models through the same broken agent. If you're picking one, none win.

#ai #llm #agents #machinelearning

12 min read

Cover image for One cheap model, one free tripwire, near-100% valid output

Torkian

Aug 11

One cheap model, one free tripwire, near-100% valid output

#ai #llm #agents #machinelearning

14 min read