Experiments or Outcomes? Probing Scientific Feasibility in Large Language Models

Create an evaluation tool that probes scientific feasibility of LLM outputs by comparing generated claims against established knowledge repositories. Developers need a way to verify scientific reasoning consistency.

Suggested repo: sci-probe

"Verify if your AI is making actual science or just science fiction."

Estimated effort: 60h