LLM作为评判者自身结果不一致？揭秘模型版本漂移与提示模糊性的根源与解法
Same outputs, same judge, two runs, two scores. The gate flickered red then green on a branch with zero code changes, and that flapping cost me more t…
看这篇论文如何将不确定性建模融入生成与决策，应对现实世界的模糊性问题。
arXiv:2606.30578v1 Announce Type: cross Abstract: With rapidly improving capabilities, Large Language Models (LLMs) are increasingly used in many comp…
多模态大模型当评委也有偏见？论文揭示文化模糊下校准失败的关键漏洞。
arXiv:2606.20676v1 Announce Type: cross Abstract: MLLM-as-a-Judge is conventionally validated by agreement with human annotations, but this metric is …
A*算法驱动多智能体，对LLM提示发起常识混淆攻击，揭示大模型语言理解的脆弱性。
arXiv:2606.01441v1 Announce Type: new Abstract: Large language models (LLMs) excel in reasoning and knowledge-intensive tasks but remain vulnerable to…
探讨模糊性在不确定性量化中对错误预测的关键影响，为机器学习可靠性研究提供新视角。
arXiv:2606.02093v1 Announce Type: cross Abstract: The task of Error Prediction, namely predicting whether a model output is correct, is commonly tackl…
用LLM结合蒙特卡洛方法建模情感轨迹与潜在歧义，为理解人际动态提供了新颖的计算视角。
arXiv:2601.03645v2 Announce Type: replace Abstract: Emotional coordination is a core property of human interaction that shapes how relational meaning …