被ICML 2026接收，提出用确定性真实值评估大模型长文本生成的不确定性，为LLM可靠性研究提供新工具。
arXiv:2607.03870v1 Announce Type: new Abstract: As LLMs generate increasingly long outputs, effective uncertainty estimation must identify errors at f…
权威综述，系统梳理LLM集成学习在文本与代码生成中的方法、挑战与未来方向。
arXiv:2503.13505v3 Announce Type: replace-cross Abstract: Generative Pretrained Transformers (GPTs) are foundational Large Language Models (LLMs) for …
用随机路径聚合可视化LLM生成中的隐藏偏见，揭秘文本背后的系统性偏差。
arXiv:2606.19344v1 Announce Type: cross Abstract: Large Language Models (LLMs) exhibit representational and syntactic biases that are difficult to eva…
零样本文本直接生成4D动态网格，颠覆传统3D建模流程
arXiv:2506.24121v3 Announce Type: replace Abstract: Large-scale, high-quality dynamic 3D (4D) assets are essential for learning physically grounded re…
从邮件中的破折号现象，揭秘LLM为何偏爱这种标点风格。
It was really uncommon pre-ai that you saw the usage of — in emails. So I wonder why all LLMs default to it so often? An example; GoDaddy: only needed…
万亿参数AI模型输出速度突破1000 tokens/s，极速文本生成开创新纪录
IT之家 6 月 12 日消息，小米 MiMo 开放平台发布公告，宣布将于 2026 年 6 月对 MiMo-V2 系列的部分模型进行正式下线处理，并推动开发者向性能更强的 V2.5 系列迁移。 IT之家注意到，此次调整涉及 mimo-v2-pro、mimo-v2-omni、mimo-v2-flas…
论文提出基于LLM并行文本生成的低延迟实时音频游戏解说系统，已被IJCAI-ECAI 2026接收。
arXiv:2606.13322v1 Announce Type: new Abstract: We present a low-latency real-time audio game commentary system that generates spoken commentary direc…
Google开源DiffusionGemma，非自回归文本模型，速度提升4倍，类似图像去噪生成。
Diffusion AI is most common in image generation, but it can make text outputs much faster.
用1800年代伦敦语料训练的老式LLM，独特的历史文本体验，可下载部署或微调
Article URL: https://crlf.link/log/entries/260525-1/ Comments URL: https://news.ycombinator.com/item?id=48487829 Points: 2 # Comments: 1
利用反事实数据实现模型感知的对比解码，让文本生成质量与真实性同步提升。
arXiv:2602.01740v3 Announce Type: replace Abstract: Video language models (Video-LLMs) are prone to hallucinations, generating plausible but ungrounde…
基于注意力机制的扩散语言模型采样器，突破传统采样效率瓶颈，推动文本生成质量提升。
arXiv:2604.08564v2 Announce Type: replace Abstract: Auto-regressive models (ARMs) have established a dominant paradigm in language modeling. However, …
用 POLARIS 引导小模型写出连贯长故事，突破规模限制，让轻量模型也能驾驭长篇叙事
arXiv:2606.04095v1 Announce Type: cross Abstract: Small open-weight models struggle at long-form creative writing: their generated stories either fall…
被ICML 2026接收，提出通过幻觉拒绝采样显著提升长文本生成的可靠性，是NLP领域的重要突破。
arXiv:2606.03628v1 Announce Type: cross Abstract: Large language models (LLMs) have achieved remarkable progress in open-ended text generation, yet th…
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…
当大模型没把握时，与其放弃回答，不如学会“模糊处理”——一种选择性抽象策略提升长文本生成可靠性
arXiv:2602.11908v3 Announce Type: replace Abstract: LLMs are widely used, yet they remain prone to factual errors that erode user trust and limit adop…
语言学感知的LLM水印新方案，无失真同时保证高检测率，平衡文本质量与版权保护。
arXiv:2606.00613v1 Announce Type: cross Abstract: Watermarking should identify language-model output without degrading quality or limiting verificatio…
带你了解全新的一致性扩散语言模型，为文本生成带来更稳定、高质量的新范式。
arXiv:2605.00161v2 Announce Type: replace Abstract: Diffusion language models (DLMs) are an attractive alternative to autoregressive models because th…
利用逆蒸馏技术加速扩散语言模型，实现更快的文本生成推理，同时保持生成质量
arXiv:2602.19066v2 Announce Type: replace Abstract: Diffusion Language Models (DLMs) have recently achieved strong results in text generation. However…
用Stacked LoRA技术将AI文本的套话频率降至零，成功骗过AI检测器。
Article URL: https://mlx-optiq.com/blog/humanizer-stacked-lora Comments URL: https://news.ycombinator.com/item?id=48353832 Points: 1 # Comments: 0
揭示多模态扩散Transformer中概念遗漏的根源，并提出基于线性探测的诊断与纠正方法。
arXiv:2605.14270v2 Announce Type: replace Abstract: Multimodal Diffusion Transformers (MM-DiTs) have achieved remarkable progress in text-to-image gen…