已验证来源
每个故事都链接到最有力的可用证据：可用的原始来源，否则明确归因的报告。
简单的英语
发生了什么、为什么重要以及值得关注的内容——无需行话。
无填料
当信号很弱时，我们不会发布任何内容，而是填充提要。
更多故事
9 故事创新
Natural Language Code Retrieval for 1C:Enterprise: An Open Benchmark and Efficient Bi-Encoder
Natural language code retrieval is a rapidly evolving task in computer science. However, the 1C:Enterprise ecosystem combines Russian syntax with highly domain-specific terminology, for which open datasets and specialized models have been virtually non-existent.arxiv.org创新
Time-Series Retrieval for Grounding Multimodal Language Models in Remaining Useful Life
Large language models (LLMs) and agentic AI systems are increasingly being explored for domain-specific maintenance and prognostics tasks, raising the question of whether they can effectively support prognostics and health management (PHM).arxiv.org创新
Nepali-English preprint: text-only AI matched multimodal model on out-of-context misinformation benchmark
A new arXiv preprint introduces NepOOC, a 1,090-pair Nepali-English benchmark for detecting misleading captions attached to authentic images. On this dataset, a text-only mBERT model matched the best tested multimodal system, while image-only models performed near chance.arxiv.org创新
Preprint finds financial AI confidence can fail when text changes
A new arXiv preprint tests financial named-entity recognition across filings, news and social media, finding that some confidence measures remain useful under domain shift while others deteriorate sharply.arxiv.org创新
对话式人工智能能否放松我们与他们的界限？
A study explores whether conversational AI can shift how majority-group members categorize and relate to Latine immigrants.arxiv.org创新
Preprint finds periodic subject changes raise judged surprise in base language models
A new arXiv preprint reports that injecting a new subject every few hundred tokens can increase language-model outputs’ judged surprise and connection. The effect did not produce integrated documents or improve the best result in an online bin-packing test, exposing limits in how long generations are evaluated.arxiv.org创新
Preprint reports declining diversity in LLM creative outputs over three years
A preliminary arXiv study finds that responses from language models have become less diverse across open-ended creativity tasks, raising questions about homogenization in human-AI creative work.arxiv.org创新
Paper reports a fine-tuning method for long-context AI with sparse attention
An arXiv preprint describes a method that trains transformer language models to work with sparse attention and reports frequent gains over models trained with exact attention.arxiv.org安全
Google Cloud Details CodeMender AI Agent for Code Vulnerability Remediation
Google Cloud presents CodeMender as an AI security agent that scans codebases, tests whether vulnerabilities can be exploited, and generates patches for developer review. The source does not describe it as a formal-verification framework or provide independent performance evidence.
cloud.google.com