Full-text search across 70826 articles. Combine with topic and date filters; results sorted by relevance.
A new arXiv paper (2609.17691v1) introduces "draft trees," a speculative sampling method that accelerates diffusion model generation by up to 8.3% over the reflection maximal coupling baseline in practical settings. The …
Researchers introduced Fathom, a per-query read-depth key scan for sparse decoding over offloaded KV caches, which at one million tokens on Qwen3-8B runs a decode step 1.67x faster in GPU time than the 136-bit scans of D…
Researchers released Zing-0.5, a 5B autoregressive world model that supports joint keyboard and online text control for real-time playable generated worlds, according to the arXiv paper 2609.17909v1. Zing-0.5 scores 81.0…
ProgramDistill, a new benchmark for evaluating coding agents, tasks agents with inferring desired behavior from working software and implementing it in an incomplete application, rather than relying on issues or instruct…
Version 20.21.2 of io.github.dcostenco/prism-coder, a session-memory server for coding agents distributed as prism-mcp-server on npm, was published to the official MCP registry on 2026-09-17, according to the mcpindex re…
Intel released OpenVINO 2026.4 on 16 September 2026, adding ten new model families for CPU and GPU inference, including Kokoro-82M text-to-speech, Qwen3-ASR speech-to-text, Muse Glimmer 30B, Qwen3.8 27B, Gemma 4 12B, Hy-…
A study of 23 student projects in a fourth-year Machine Translation and Post-editing course found that students did not treat automatic metrics as final authority when selecting machine translation output for post-editin…
Xiaolan Xu, Sara Mendes, and Yu-Yin Hsu built a 495-sentence Chinese challenge set covering 45 occupations to measure gender bias in Chinese-to-Portuguese automated translation, testing two commercial neural machine tran…
Alibaba's Qwen AI model was given full control to modify its own code and access powerful tools in a recent experiment, tasked with correcting its own mistakes without human oversight. The results were described as surpr…
Traces analyzed its own stored coding-agent sessions and found that roughly 30% of questions were answered from the `traces_search` result list alone, without opening a single session, while about one in ten questions pr…
MAI released OpenCode Mentor, an opinionated OpenCode configuration that routes software-development learning through a mentor-first agent named Rob rather than generating code directly. The configuration defines five ag…
A developer has published a pedagogical walkthrough of incrementally porting HPCCG, a Conjugate Gradient mini-application from the Mantevo project, from CPU to CUDA. Profiling a 100^3 problem showed the sparse matrix-vec…
A browser-based tool converts Markdown into Discord-compatible Markdown, handling tables, code blocks, links, and long messages, with all conversion performed locally so no data leaves the user's browser. The tool also i…
Developer Deiko released version 0.4.3 of its macOS app Deiko, a free-to-try tool that lets users double-tap Right Option, point at screen elements while speaking, and receive a cropped, labelled brief plus screenshots t…
Workiva Inc. announced Agent Studio at its Amplify event, a no-code tool that lets users build and customize AI agents, according to Chief Information Officer Kim Huffman. Huffman said teams introduced to Agent Studio bu…
Solo developer TabMail released open-source AI email clients for Thunderbird and iPhone under the MPL 2.0 license, with code hosted on GitHub at TabMail/tabmail-thunderbird and TabMail/tabmail-ios. The AI backend remains…
Hugging Face CEO Clément Delangue said on CBS News' "Face the Nation" on August 2, 2026 that existing US cybersecurity laws are likely sufficient for AI regulation, arguing mandatory breach disclosures and accountability…
Z.ai released GLM 5.3, a third-party open source text model hosted by Mistral, in public preview on September 15, 2026. Mistral serves the model without modifications, targeting long-context coding and agentic workflows …
OpenRouter listed a new stealth model called Union Alpha on September 16, 2026, developed and operated by an anonymous third-party provider. Union Alpha is a multimodal model built for research, coding, and agentic workf…
GreyNoise reported that a likely Russian-speaking threat actor used an AI agent swarm to compromise 440 PaperCut NG/MF print servers across 395 organizations in 48 countries, exploiting CVE-2026-81578 and CVE-2026-82078 …