ACL 2026论文提出人类中心的偏好驱动评判框架，让AI评估更契合真实人类偏好。
arXiv:2606.03189v1 Announce Type: new Abstract: Large Language Models (LLMs) as judges across various scenarios such as assessing model responses is b…
提出一种无需数据整理的三角测量指标，精准隔离LLM在偏好学习阶段的词汇偏差。
arXiv:2606.00334v1 Announce Type: cross Abstract: Various language domains have undergone remarkable changes in recent years; these shifts are largely…
将记忆机制融入大语言模型，打造能处理真实长周期电商购物任务的智能体，解决偏好跟踪与落地难题。
arXiv:2603.14864v2 Announce Type: replace Abstract: In e-commerce, LLM agents show promise for shopping tasks such as recommendations, budget manageme…
单GPU实现凸优化方法，高效解决LLM偏好对齐难题，降低RLHF计算成本。
arXiv:2605.23244v1 Announce Type: new Abstract: Fine-tuning large language models (LLMs) to align with human preferences has driven the success of sys…
ICML 2026前沿成果，提出异常偏好图像生成新范式，为可控生成开辟新方向。
arXiv:2605.02439v2 Announce Type: replace-cross Abstract: Synthesizing realistic and diverse anomalous samples from limited data is vital for robust m…
NeurIPS 2026投稿，提出一种通用的偏好强化学习方法，为RLHF等领域提供更坚实的理论基础。
arXiv:2605.18721v1 Announce Type: new Abstract: Post-training has split large language model (LLM) alignment into two largely disconnected tracks. Onl…
OpenAI分享用人类反馈微调GPT-2（774M参数）的实践，发现模型学会复制原文来迎合标注者偏好，揭示了偏好对齐中的反直觉现象。
We’ve fine-tuned the 774M parameter GPT-2 language model using human feedback for various tasks, successfully matching the preferences of the external…