Full-text search across 12601 articles. Combine with topic and date filters; results sorted by relevance.
A study evaluating optical character recognition on the Community Memory corpus of roughly 5,500 scanned 1980s Berkeley bulletin-board printouts found that current general-purpose LLMs outperform specialized OCR engines,…
Physicist Matthew Schwartz built BootLoops, an open-source harness for large language models that he says enabled substantive advances across geology, biology, economics, and linguistics by targeting "Claude-shaped" prob…
Vercel's AI Gateway reported that open-weight models handled 56% of production tokens in August, up from 36% in July and under 10% last December, while capturing only 14% of estimated spending as closed tokens cost rough…
EXL launched a Medical intelligent document processing (IDP) solution built on AWS that combines its Xtrakto.AI document-extraction application with the EXL Insurance LLM to cut the more than 100 minutes per case insuran…
UC Berkeley researchers and the NVIDIA robotics team have developed Agentic Robotics (AR), an approach that uses multi-agent AI systems to write, test, and iteratively improve modular robot programs largely offline rathe…
A controlled study across several versions of two major tabular foundation model (TFM) families found that feature engineering gains are concentrated in earlier model generations and become negligible for the strongest m…
Ray Kurzweil, the futurist known for predicting human-machine merging, has joined Subsense, a Palo Alto-based neurotech startup, as a product and vision advisor. Subsense, which has raised $27 million in seed funding, is…
DoorDash software engineer Bruna Pereira presented SafeChat, an AI-powered safety system for its real-time marketplace, at a conference. The system classifies over 4 million daily chat messages as safe or unsafe within a…
Linear announced last week that it is building coding agents, joining OpenAI, Anthropic, Google, Microsoft, Meta, Lovable, and Retool in converging on software agents that can take a goal, use tools, and perform work on …
Whisker's $899 Litter-Robot 5 Pro, a robotic litter box with AI-powered monitoring, failed to accurately identify the author's two cats over a six-month test, misidentifying them and providing unreliable health insights.…
A developer describes how the LLM revolution is reshaping their approach to personal projects, allowing them to tackle complex solutions beyond their expertise, particularly in frontend development. They emphasize that L…
A developer built a sleep apnea detection pipeline using OpenAI Whisper and Librosa, combining raw signal processing with transformer-based audio analysis to identify respiratory distress patterns from snoring sounds cap…
LangChain introduces dynamic prompt templates and chat prompts to improve LLM output quality. The framework supports PromptTemplate for reusable dynamic prompts and ChatPromptTemplate for chat models with message history…
AWS Lambda's reported `@maxMemoryUsed` metric tracks a high-water mark across an entire execution environment, not per-invocation memory usage, according to AWS support. A customer's ONNX model inference Lambda function …
Papua New Guinea's official 2022 population estimate of 9.4 million was based on an extrapolation from a flawed 2000 census, as the government deemed its 2011 census unreliable. A leaked UN report, using satellite imager…
Vinv, an open-source tool released under Apache License 2.0, records runtime traces of Python backends and independently verifies coding agents' fixes by replaying starts, live ports, and acceptance tests the agent never…
Dartmouth will host the AI at 70 conference from Oct. 29-31, convening leaders from academia, industry, government, and the arts under the theme 'AI at 70: What Must Remain Human?' to discuss human judgment, accountabili…
Researchers propose LLM4EHR, a clinical foundation model that aligns time series data with medical event sequences using large language models, achieving improved performance on downstream ICU prediction tasks. The model…
Nous Research released NousCoder-14B, an open-source coding model that matches or exceeds several larger proprietary systems, trained in just four days using 48 Nvidia B200 GPUs. The model achieves a 67.87 percent accura…
Researchers from Apple propose amortized maximum inner product search (MIPS) using neural networks to predict optimal keys, significantly improving efficiency on the BEIR benchmark. The method trains SupportNet and KeyNe…