Full-text search across 12648 articles. Combine with topic and date filters; results sorted by relevance.
A developer built a provider-agnostic RAG pipeline using hexagonal architecture, decoupling the application from specific LLMs, vector databases, and document-processing frameworks through lazy imports and abstract inter…
Reactor and the Amazon Neuron Science team developed a kernel-centric optimization strategy that enables real-time, low-latency video generation on AWS Trainium chips, according to Reactor co-founder and CTO Bryce Schmid…
Perplexity Research published a post-training study that trains a model inside Perplexity Computer on real user sessions, including failed ones, pairing rejection sampling fine-tuning with hint-guided self-distillation. …
A research team led by the Leibniz Institute for Food Systems Biology at the Technical University of Munich developed an AI method that predicts peptide bitterness and designs new bitter-tasting peptides from scratch, wi…
Researchers introduced UltraBench 2, a comprehensive benchmark for evaluating vision foundation models on ultrasound image analysis, with wide anatomical and task coverage and a focus on standardization, reproducibility,…
Researchers Samuel McCandlish and colleagues submitted "Scaling Laws for Neural Language Models" to arXiv on 23 January 2020, showing that language model cross-entropy loss falls as a power-law with model size, dataset s…
Vansh Deo and team QueryCrew built Zyg0s, an autonomous, explainable fraud investigation agent that fuses seven neuro-symbolic agents with TigerGraph Savanna Cloud, Model Context Protocol, and Groq to traverse entity net…
DuckDB can query Hugging Face Hub datasets directly over its hf:// protocol without downloading them, an integration added in DuckDB v0.10.3, released May 22, 2024, and announced in May 2024. The scheme, built on DuckDB'…
Zalando has open-sourced the Agentic Identity Broker, an MIT-licensed component that lets AI agents act on behalf of users across third-party systems without ever holding the users' provider tokens. The broker handles us…
BottleCap AI released ThinkingCap-Qwen3.8-27B, a fine-tune of Qwen3.8-27B that cuts thinking tokens by an average of 37.2% across 12 benchmarks while macro-average accuracy falls 0.86 percentage points, from 86.65% to 85…
Anthropic CEO Dario Amodei announced that the company's Claude AI discovered a previously unknown enzyme system, named ART, in the DNA of Staphylococcus phages, using roughly 950 AI agents and 210 million tokens to surve…
Fresh Jots launched a tamper-evident note service that hashes and anchors every entry to the Bitcoin blockchain, letting users, scripts and AI agents prove a log existed at a certain time and was not altered. The service…
A developer is building Soil Doctor, an agricultural advisory system whose retrieval-augmented generation pipeline combines semantic embeddings, BM25 lexical search, and cross-encoder reranking over a prototype knowledge…
DX released CAFE(S), a framework for evaluating the quality of context assembled for AI systems, published in ACM Queue and co-authored by Max Kanat-Alexander (Capital One), Eirini Kalliamvakou (DX), Margaret-Anne Storey…
Google announced that AlloyDB now delivers PostgreSQL for agents in preview, provisioning sandboxed database instances in seconds with full workload isolation from production. The agentic database architecture scales to …
Liquid AI released an experimental 279.5-million-parameter draft model, DSpark, for its LFM2.5-VL-3B vision-language model on September 24th, which the company says cuts decoding time by up to 3.13x on an M5 Max without …
Caterpillar's chief digital officer outlined five strategies the company used to build the Cat AI Assistant, an AI agent deployed in under a year to help address a 37,000-per-year shortage of skilled technicians in the U…
A technical blueprint examines the trade-off AI coding assistants make between generation speed and accuracy, arguing that models must explicitly encode contextual information such as code structure, module dependencies,…
Huawei launched the Atlas 960E SuperPoD AI cluster, which packs 4,096 Ascend NPUs with proprietary optical interconnects, alongside the OceanStor M900 context memory storage built to offload large Key-Value caches, accor…
Medical AI developers should select a single first regulatory claim — such as measuring a radiologist-selected pulmonary nodule on non-contrast chest CT — before funding validation, because each of five candidate claims …