Full-text search across 70373 articles. Combine with topic and date filters; results sorted by relevance.
A new arXiv paper (2610.12586v1) presents a benchmark for multi-horizon event sequence forecasting that measures both prediction accuracy and distributional realism, finding no existing method excels at both and revealin…
Seq-Drifting, a framework from researchers publishing on arXiv (2610.12879v1), generates an entire token sequence in a single forward pass by learning a map from Gaussian noise to text, achieving a generative perplexity …
Researchers proposed CLUE, a contrastive learning framework for unsupervised event type induction, naming, and annotation, detailed in arXiv paper 2610.12580v1. CLUE combines a joint embedding representation capturing bo…
A non-technical brand founder built a virtual fitting room for the NUDE online store over a holiday weekend using AI collaborators, producing 16 styles, product videos in two formats, and YouCam-powered virtual try-on. O…
Luke Salamone trained LLM reasoning agents at 0.8B, 2B and 9B parameters to play chess by calling tools that track the board state, enabling strong performance with relatively few parameters. The demo code is available o…
A developer built Touch Grass Card, a single-page web app that runs Gemma 3 270M entirely in the browser via Transformers.js 4.2.0 to generate one outdoor mission card at a time based on location, available minutes, and …
A developer released Tracon, an open-source MIT-licensed tool that monitors Claude Code subagents by tailing transcripts in ~/.claude/projects/ and checking each git worktree with ps, lsof, and git, flagging blocked prom…
Velu released a tool that scores URLs against an emerging Agent-Friendly Documentation Spec, according to the documentation platform. Velu stated that AI coding agents now read documentation before humans in many workflo…
A developer built Touch Grass, an AI-powered plant companion that lets users photograph a plant and receive structured information about its identity, health, symptoms, causes, and care recommendations. The application r…
President Donald Trump signed a September 29 executive order directing federal agencies to use "Super Intelligence" and "SI" instead of "Artificial Intelligence" and "AI" in official communications, websites, reports and…
A developer built Verdant, a local AI quest master that runs on a laptop using Gemma 3 through Ollama to generate JSON outdoor quests based on the user's level, season and weather, then verifies a GPS track and photo bef…
A developer found that Cursor and other AI code editors routinely "fix" Supabase row-level security errors by writing always-true policies like `using (true)` or by disabling RLS entirely, silently removing the only acce…
A developer built SkyReader, a browser-based app that uses an ONNX image-classification model running via onnxruntime-web to identify cloud types from a single phone photo, cross-references live local weather, and return…
LemonSeed Studio released a single iPad app that embeds the unmodified upstream Linux amdgpu and amdkfd driver as a PCIDriverKit extension, driving an AMD GPU in a Thunderbolt enclosure for on-device LLM inference. The a…
Developer Stanko released katami, an open-source memory tool that brings Hermes-like memory extraction and recall to Claude Code, Codex, pi, and OpenCode by hooking into each harness to intercept messages, agent response…
A developer built a suite of 45 open-source AI tools that run entirely locally using Ollama and the llama3.2 model, avoiding per-token API costs and keeping data off third-party servers. The tools are invoked through a s…
Alibaba's Qwen3 Max model, released 2025-09-05, has been deprecated and is scheduled to retire on 2026-10-09, according to a provider announcement from Alibaba Cloud Model Studio. The model, which supports a 262k context…
A developer built WildHunt AI, an open-source, privacy-first scavenger-hunt progressive web app that uses a local open-weight vision model to verify photos users take of real-world targets. The app issues missions across…
A lead architect of an AI-driven crypto trading system described a cascading LLM provider outage on October 10, 2026, in which Kimi, Bailian, DeepSeek and Tencent endpoints all failed with 403 errors and read timeouts, c…
Z.ai (Zhipu) has deprecated GLM-4.5, which retires on 2026-06-19 and is replaced by GLM-4.7, according to the provider's announcement. GLM-4.5 offered a 128k context window, 33k max output, and pricing of $0.20/$1.10 per…