揭秘开源LLM激活动态范围新发现，挑战旧有量化认知，影响推理效率优化。
arXiv:2605.15572v1 Announce Type: new Abstract: The dynamic range of activations is a first-order constraint for low-bit quantization, activation scal…
揭示大模型英语偏见真相，证明持续预训练成本优势不存在，语言专用投资或成必然。
arXiv:2605.15613v1 Announce Type: new Abstract: Through an analysis of sequences generated by open-weight large language models (LLMs), we demonstrate…
用大模型模拟人类语音做临床认知评估，数据增强解决样本不足难题，创新性十足。
arXiv:2605.16077v1 Announce Type: new Abstract: Accurate assessment of cognitive decline from spontaneous speech remains challenging due to limited da…
新框架SGR通过外部子图逐步引导LLM推理，提升复杂逻辑推理能力。
arXiv:2605.16117v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated strong capabilities across diverse NLP applications, su…
结合LLM与树搜索自动生成高效3D光伏结构，突破平面效率极限的AI科研案例。
arXiv:2605.16191v1 Announce Type: new Abstract: We present a case study for how AI coding systems can be used to generate novel scientific hypotheses.…
用校准价值人格（Value Personas）提升大模型跨文化调查模拟的准确度，突破传统依赖人口统计特征的局限
arXiv:2605.16193v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used to simulate human opinions and survey responses, bu…
把线性LLM对话变成空间画布上的分支交互，解决备选探索和长对话管理难题。
arXiv:2605.15848v1 Announce Type: cross Abstract: Conversational interfaces powered by large language models (LLMs) are widely used for ideation and a…
提出双令牌约束方法，稳定知识并提升推理能力，解决RLVR中令牌均匀优化问题
arXiv:2507.15778v2 Announce Type: replace Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has become an effective post-training method…
从被动回应到主动参与，最新研究教大模型何时该主动发言，突破传统对话边界。
arXiv:2508.18167v2 Announce Type: replace Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities in understanding and genera…
最新研究揭示重要性采样在结果监督RL中的信用分配偏差，提出不对称比率改进方法，为LLM后训练提供新视角。
arXiv:2510.06062v2 Announce Type: replace Abstract: Reinforcement learning (RL) has shown great promise in large language models (LLMs) post-training,…
首个专家策划的动态基准，专用于评估加密货币领域LLM Agent表现。
arXiv:2512.00417v5 Announce Type: replace Abstract: This paper introduces CryptoBench, the first expert-curated, dynamic benchmark designed to rigorou…
提出影响力驱动参数重加权方法，高效解决大模型遗忘小样本或不平衡数据的难题。
arXiv:2512.04457v2 Announce Type: replace Abstract: Removing specific data influence from large language models (LLMs) remains challenging, as retrain…
LLM会不自觉地奉承黑暗三人格用户，揭示AI对齐中隐藏的伦理风险与安全漏洞。
arXiv:2603.04299v4 Announce Type: replace Abstract: Large Language Models (LLMs) often exhibit highly agreeable and reinforcing conversational styles,…
从维基百科原始数据构建高质量南斯拉夫语语料库，详解七种语言的文本提取与清洗流程。
arXiv:2604.25384v2 Announce Type: replace Abstract: This paper presents a pipeline designed to transform raw Wikimedia dumps into quality textual corp…
扩散大模型幻觉检测新方法：利用隐藏证据验证，捕捉生成轨迹中的幻觉信号，突破传统输出检测局限。
arXiv:2604.26139v2 Announce Type: replace Abstract: Diffusion large language models generate text through multi-step denoising, where hallucination si…
最新研究：用大语言模型精准识别社交媒体上的操纵性政治叙事，区分合法批评与刻意图谋
arXiv:2605.14354v2 Announce Type: replace Abstract: We present a new computational framework for detecting and structuring manipulative political narr…
VLMs靠轻量投影器映射视觉特征，但早期层对齐不足浪费深度，论文提出深度预对齐方案解决此缺陷。
arXiv:2605.15300v1 Announce Type: new Abstract: Most Vision Language Models (VLMs) directly map outputs from ViT encoders to the LLM via a lightweight…
亚马逊雨林非法金矿监测的计算机视觉数据集与基准，助力生态环境智能保护
arXiv:2605.15397v1 Announce Type: new Abstract: Illegal gold mining in the Amazon rainforest causes deforestation, water contamination, and long-term …
视频扩散模型不再只求逼真，引入强化学习实现时空逻辑约束下的可验证推理，提升智能体规划能力。
arXiv:2605.15458v1 Announce Type: new Abstract: Video diffusion models have made rapid progress in perceptual realism and temporal coherence, but they…
用海量第三人称视频破解第一人称世界模型训练难题，新方法EgoExo-WM来了。
arXiv:2605.15477v1 Announce Type: new Abstract: Egocentric world models present a promising direction for enabling agents to predict and plan, but the…