When language models revise a decision after another agent disagrees, neither reversal counts nor final accuracy shows whether they are fixing mistakes or ab…
机构:Georgia Tech
来源:arXiv 2611.16694 | AI4Papers 论文推荐平台