Training-free Truthfulness Detection via Value Vectors in LLMs

Re-assign community