x

LESSWRONG

LW

Shahriar Golchin — LessWrong

Shahriar Golchin

Shahriar Golchin

Message

I am an AI Researcher at Labelbox, working on model evaluation and AI safety.

I earned my PhD in Computer Science (AI/ML) from the University of Arizona. My research interests center on Large Language Models (LLMs), focusing on designing datasets/tasks that adversarially stress-test alignment. I am...

4

1

1mo

Shahriar Golchin

I am an AI Researcher at Labelbox, working on model evaluation and AI safety.

I earned my PhD in Computer Science (AI/ML) from the University of Arizona. My research interests center on Large Language Models (LLMs), focusing on designing datasets/tasks that adversarially stress-test alignment. I am...

The AI Safety Illusion: Why Current Safety Datasets Fool Us on Model Safety

AI models are increasingly trained to be “safe,” meaning they refuse harmful requests. But what does it truly mean for a model to be safe? Today, this is typically assessed using safety benchmarks: curated collections of adversarial prompts designed to test whether a model will refuse dangerous behavior. If the...