LIFEBENCH: Evaluating Length Instruction Following in Large Language Models | Read Paper on Bytez