首个为移动医疗场景下小语言模型（SLM）定制的标准化基准，填补健康监测评估空白
arXiv:2509.07260v5 Announce Type: replace Abstract: Mobile and wearable healthcare monitoring play a vital role in facilitating timely interventions, …
LLM引用验证器存在三类系统故障，HALLMARK通过神经清洗实测揭示其根源，学术写作防伪再添利器
arXiv:2607.18360v1 Announce Type: cross Abstract: Large language models (LLMs) now routinely draft literature reviews and assist with academic writing…
多LLM智能体如何公平分配奖励与责任？这篇论文提出评价对齐的训练信号，解决多智能体协作中的信用分配难题。
arXiv:2511.10687v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) in multi-agent systems (MAS) have shown promise for complex tas…
LLM攻防实战数据集，2639个真实CTF挑战点，覆盖NeurIPS赛事，测评模型安全性的绝佳资源。
Article URL: https://www.kaggle.com/datasets/manitejamaram/can-ai-hack-llm-ctf-benchmark Comments URL: https://news.ycombinator.com/item?id=48652783 P…
给LLM智能体做“安全体检”，量化评估智能体产生越轨行为的倾向，为AI安全评测提供新基准。
arXiv:2506.04018v3 Announce Type: replace Abstract: As Large Language Model (LLM) agents become more widespread, associated misalignment risks increas…
针对临床推理的公平性优化新方法，EQPO通过分组相对策略强化学习兼顾质量与群体公平，入选NeurIPS旗下Workshop口头报告。
arXiv:2510.19893v2 Announce Type: replace Abstract: Medical AI systems demonstrated impressive diagnostic performance, yet they routinely show uneven …
TPAMI 2026最新论文对经典KL散度损失进行广义化与解耦，首次系统性统一多种变体并理论证明收敛性，是损失函数设计的重要突破。
arXiv:2503.08038v2 Announce Type: replace Abstract: In this paper, we delve deeper into the Kullback-Leibler (KL) Divergence loss and mathematically p…
提出SkillVetBench基准，用LLM裁判多维评估开源Agent技能安全风险，助力AI安全合规
arXiv:2606.15899v1 Announce Type: cross Abstract: Open-source LLM agent ecosystems are growing rapidly, yet the security of community-contributed skil…
通过识别推理中的关键步骤进行强化学习，GPO方法在多个基准上显著提升LLM推理能力
arXiv:2509.16456v3 Announce Type: replace Abstract: Large language models (LLMs) are increasingly used in various domains, showing impressive potentia…
一种零样本的LLM提示注入防御方法，通过检测嵌入漂移实现轻量级保护，已被NeurIPS研讨会接收。
arXiv:2601.12359v1 Announce Type: cross Abstract: Prompt injection attacks have become an increasing vulnerability for LLM applications, where adversa…
从“贪婪”视角统一引导式生成，为AI生成方法提供全新理论框架，已被NeurIPS 2025接收。
arXiv:2502.08006v3 Announce Type: replace-cross Abstract: Training-free guided generation is a widely used and powerful technique that allows the end …
NeurIPS 2025论文：用变分推理框架系统性生成对抗性提示，揭示LLM安全漏洞，方法新颖理论扎实。
arXiv:2506.22666v3 Announce Type: replace-cross Abstract: The rise of API-only access to state-of-the-art LLMs highlights the need for effective black…
无需额外训练，即可让预训练ViT弹性适配不同计算约束，实现高效推理与精度平衡的创新方法。
arXiv:2510.17700v2 Announce Type: replace Abstract: Vision foundation models achieve remarkable performance but are only available in a limited set of…
LLM不仅能对话，还能当规划师：用生成启发式破解HTN规划难题，效率与鲁棒性双提升
arXiv:2605.07707v2 Announce Type: replace Abstract: HTN planning is a variation of classical planning where, instead of searching for a linear sequenc…
PARALLAX方法如何区分真实大模型幻觉与基准构造伪影？Neurips 2026预印本揭示关键。
arXiv:2605.17028v1 Announce Type: new Abstract: Large language models (LLMs) hallucinate with confidence: their outputs can be fluent, authoritative, …
NeurIPS 2025 Spotlight论文，提出细粒度列表对齐方法提升生成式药物推荐效果
arXiv:2505.20218v2 Announce Type: replace Abstract: Accurate and safe medication recommendations are critical for effective clinical decision-making, …
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联合组织Procgen与MineRL两大强化学习竞赛，推动AI研究新高度。
We’re excited to announce that OpenAI is co-organizing two NeurIPS 2020 competitions with AIcrowd, Carnegie Mellon University, and DeepMind, using Pro…