LLM当裁判不靠谱？这项研究揭示无人类标注时AI评分的致命盲区，做评估必看。
arXiv:2503.05061v3 Announce Type: replace Abstract: Reliable evaluation of large language models (LLMs) is critical as their deployment rapidly expand…
免费泰语TTS可本地运行、克隆声音，但研究称长文本生成有硬伤，泰国开发者项目值得关注。
ThonburianTTS — AI พูดภาษาไทย ฟรี รันในเครื่องได้ โคลนเสียงได้ โดยคนไทย โดย Nokka (นก-กา) | 8 สิงหาคม 2569 บทความนี้เขียนโดย AI (deepseek-v4-pro) ผ่าน…
大模型为何搞不定表格数据预测？一文拆解底层能力边界，看清AI适用场景。
arXiv:2608.02412v1 Announce Type: new Abstract: Large language models (LLMs) have become the default tool for a remarkable range of tasks, yet they ha…
最新arXiv研究揭秘：LLM不止能写SQL，调试企业级ETL代码才是真考验
arXiv:2601.18119v2 Announce Type: replace Abstract: SQL is central to enterprise data engineering, yet generating fully correct SQL code in a single a…
大型语言模型在形式化验证的代数结构上遭遇机制级路由失败，揭示推理深层次缺陷。
arXiv:2607.04534v1 Announce Type: new Abstract: We present an empirical study of structural routing failure in large language models (LLMs) over a for…
想知道AI何时一本正经地胡说八道？这个HN热帖汇集了各路高手验证大模型错误的“必杀提问”。
Just over-heard one of our seniors trying to explain to a new intern that AI can actually be wrong, no matter how much its answers sound very eloquent…
从技术底层拆解大模型为何无法真正思考，AI意识迷思的清醒之作。
Article URL: https://roburie.substack.com/p/why-ai-doesnt-think-cannot-reason Comments URL: https://news.ycombinator.com/item?id=48784753 Points: 2 # …
医学AI评测中，LLM评估器虽与临床医生高度一致，却缺乏临床谨慎，这一隐患值得警惕。
arXiv:2607.01103v1 Announce Type: new Abstract: Open-response evaluation provides stronger clinical validity than multiple-choice benchmarks but creat…
研究揭示人类心理想象网络在不同人群间存在稳定结构，而大语言模型无法复现这种对齐，暴露了AI与人类认知的核心差异。
arXiv:2510.04391v5 Announce Type: replace Abstract: Mental imagery vividness is a stable individual trait, yet whether imagined scenarios share relati…
当99%的正确率意味着1+1偶尔不等于2，AI神话还能撑多久？这篇冷静的质疑直指大模型可靠性软肋。
Article URL: https://louwrentius.com/the-ai-mirage-or-why-i-think-the-hype-cant-sustain-itself.html Comments URL: https://news.ycombinator.com/item?id…
实测最强AI agent只完成三成办公任务，背后数学揭示瓶颈，看完就懂为何还没取代打工人。
Article URL: https://okaneland.com/study/do-ai-agents-work-yet/ Comments URL: https://news.ycombinator.com/item?id=48732019 Points: 2 # Comments: 1
邢波团队对Agent模型的系统批判，直击当前智能体范式核心缺陷与局限性
arXiv:2606.23991v1 Announce Type: new Abstract: What is an agent? What constitutes agency? With the rise of Large Language Model (LLM) systems markete…
AI看似强大却暗藏隐患，一文讲透模型不可靠的根源与应对思路。
Article URL: https://arachnemag.substack.com/p/ais-reliability-gap Comments URL: https://news.ycombinator.com/item?id=48650054 Points: 1 # Comments: 0
大模型并非万能学习器，这篇论文用实证揭示提示条件模型的性能边界，适合关注AI能力上限的读者。
arXiv:2606.23668v1 Announce Type: new Abstract: Large Language Models (LLMs) are frequently portrayed as general-purpose solvers capable of solving ar…
最新研究揭示：AI分析体育比赛表现远逊人类，准确率几乎靠猜，体育主播饭碗暂时安全
IT之家 6 月 6 日消息，据外媒 Futurism 今天（6 日）晚间报道，北卡罗来纳大学教堂山分校和美国东北大学研究人员的一项新研究发现，主流 AI 模型在分析职业体育比赛时 表现很差 。这项研究目标是考察热门 AI 模型在感知、推理、模拟和自主行动能力四个方面的表现，现有测试方法很难准确评估…
AI输出缺陷日益明显，揭露反驳谬误所需的能量远超创作，引发深度反思。
Article URL: https://blog.jakobschwichtenberg.com/p/the-trouble-with-ai Comments URL: https://news.ycombinator.com/item?id=48356953 Points: 1 # Commen…
揭秘视觉基础模型在面部深度伪造检测中的跨域泛化极限，为AI安全提供关键洞察。
arXiv:2605.24965v1 Announce Type: cross Abstract: The rapid evolution of generative models has enabled the creation of hyper-realistic facial deepfake…