Full-text search across 7151 articles. Combine with topic and date filters; results sorted by relevance.
AI investment is on pace to top a trillion dollars a year, diverting capital, industrial real estate, electricity, and labor from housing, manufacturing, and other sectors, according to a New York Times piece by Jennifer…
A new analysis explores how weight quantization affects model performance and sparse autoencoder (SAE) reconstruction of the residual stream, using fraction of variance unexplained (FVU) as a key metric. The study, inspi…
Researchers introduced AutoBackSwap, a method that reduces AI image classifiers' reliance on spurious background correlations by automatically swapping backgrounds in training data. The approach improves model robustness…
A new unified AI model integrates textual reasoning and graph processing into a single masked diffusion language model, outperforming graph neural networks, graph transformers, and LLM-based baselines by up to 3.9 points…
Researchers introduced LINet, a multi-stream neural network for RGB-D scene classification that outperforms existing methods through a continuous integration strategy. LINet uses a novel Linear Integration Convolution op…
Anthropic has introduced prompt caching for its Claude API, allowing developers to resume from specific prefixes in prompts to reduce processing time and costs. The feature supports automatic and explicit caching with a …
Researchers introduced Hierarchical Global Attention (HGA), a method that reduces GPU memory usage in long-context transformers by using hierarchical routing, enabling a 64K-token context on an RTX 5090 GPU with 32GB mem…
Researchers introduced MCO-PDE, a competitive optimization framework that discovers shared partial differential equations from multi-source datasets by fusing as few as 50 observations per dataset. The framework uses neu…
Researchers introduced Symbolic Mechanistic Data Attribution (SMDA), a framework that links training data to high-level model behaviors, tested on Llama-3.2-3B-Instruct. SMDA revealed systematic safety gaps and biases, o…
A glossary defines terms used in technical writing and AI, including agentic workflows, LLMs, MCP, RAG, and vibe coding, along with coined terms like devling, product truth, documentation theater, the Ferri Paradox, NVE …
Andrzej Janik released ZLUDA v6 on June 29, adding 32-bit PhysX support and AI workload improvements, but announced commercial funding has ended for the third time, returning the project to a weekend hobby. The drop-in C…
Britain's AI sector has more than tripled in value to £41bn in a year, driven by firms like Nscale (£11.3bn) and ElevenLabs (£8.5bn), as the UK cements its position as Europe's leading startup hub with 70 unicorns worth …
A report from the Business and Policy Institute documents 21 campaigns by the Party for Socialism and Liberation (PSL) across 14 states that blocked or delayed $23.6 billion in AI infrastructure investment. The PSL is li…
Ford admitted its AI-driven quality push fell short and hired over 350 veteran engineers over three years to fix costly quality problems. The automaker concluded AI alone could not deliver expected improvements, but comb…
Transformer models develop outlier channels—feature dimensions with unusually large values in weights and activations—due to the softmax normalization in attention layers, which forces tokens to assign importance even wh…
Databricks co-founder Reynold Xin explains the limitations of traditional monolithic OLTP databases and introduces Lakebase, a serverless Postgres database that externalizes storage and compute. The architecture separate…
Animal advocates face a definitional challenge in aligning AI with animal welfare, as they must determine how much moral weight to assign to different animals and which outcomes to prioritize. The arrival of superintelli…
Anthropic launched Claude Science, an AI workbench for scientists that integrates over 60 databases and pre-built toolkits, aiming to streamline computational research without introducing a new model. The platform includ…
Anthropic's Mythos AI model was reinstated by the Trump administration for limited release to over 100 U.S. institutions after a two-week block, while OpenAI's GPT-5.6 lineup faced similar restrictions. New data from Acc…
IBS Software used Amazon Bedrock's managed distillation capabilities to build a bilingual named entity recognition system for cargo logistics, achieving 95.085% F1-Score accuracy while reducing operational costs by 14x. …