Large language models (LLMs) increasingly participate in scientific evaluation, both as automated reviewers and as assistants to human reviewers. As model-ge…
机构:Maryland
来源:arXiv 2609.20942 | AI4Papers 论文推荐平台