多角色评分标准生成，让LLM裁判与奖励模型更稳健、更对齐人类偏好。
arXiv:2607.01830v1 Announce Type: new Abstract: Reliable reward and preference signals are critical for evaluating and optimizing large language model…
新研究揭示LLM裁判在代理任务中打分不可靠，瓶颈与改进思路值得关注
arXiv:2606.29920v1 Announce Type: new Abstract: Rubric-based scoring has become a widely used paradigm in model evaluation, typically with LLM-as-a-Ju…
揭秘LLM作为评判者时的位置偏差，点式与对式评估究竟何者更准确？
arXiv:2602.02219v2 Announce Type: replace Abstract: Large language models are widely employed as evaluators, a paradigm commonly referred to as LLM-as…
为开放法律任务设计的评分标准引导诊断基准，提升法律AI系统评估的可靠性和可解释性。
arXiv:2606.09389v1 Announce Type: new Abstract: As large language models (LLMs) are increasingly applied to real-world legal tasks, evaluating the rel…
提出专家评分标准解决RLVR中复杂约束问题，为强化学习奖励设计提供新范式
arXiv:2606.09118v1 Announce Type: new Abstract: As LLM capabilities advance rapidly, the evaluation methods used to assess them increasingly lag behin…
用BART模型自动评估CS1的C++编程作业，基于评分标准实现高效、公正的代码打分。
arXiv:2606.03814v1 Announce Type: new Abstract: This paper investigates rubric-aware, multitask fine-tuning of transformer models for automated gradin…
提出偏好感知评分标准学习方法，让AI评估更贴合个人标准，解决评测千人千面难题
arXiv:2605.31545v1 Announce Type: new Abstract: As Large Language Models (LLMs) evolve from general-purpose assistants to user-centric agents, persona…
AI开始给人类打分，Claude评分标准曝光，优秀人类仅得7.5分，揭示人机交互新维度
反转了反转了，过去我们给 AI 跑分，今天 Claude 开始反手给人类打分！它会通过 11 个指标来分析你和它的历史对话，判断你使用 AI 的水平高低。在 AI 眼里，你是高手还是萌新？ 倒反天罡，AI 开始给人类打分了！ 搞出这个功能的，还是那个抽象的 Anthropic。 这个功能虽然目前还在…
用评分标准做增量训练，让大模型在开放式医疗对话中也能高效对齐，突破传统RL局限。
arXiv:2510.15859v4 Announce Type: replace-cross Abstract: Reinforcement learning (RL) has driven recent breakthroughs in large language models (LLMs),…
提出记忆增强的评分标准改进系统，提升基于评分标准的强化学习效果。
arXiv:2605.18592v1 Announce Type: new Abstract: Rubric-based reward shaping is an effective method for fine-tuning LLMs via RL, where structured rubri…