Experiments or Outcomes? Probing Scientific Feasibility in Large Language Models | Read Paper on Bytez