Full-text search across 70296 articles. Combine with topic and date filters; results sorted by relevance.
UC Berkeley and NVIDIA released open-source tools that let teams of AI coding agents write, simulate and refine readable robot programs, according to The Robot Report. The release targets industrial robot programming wit…
A researcher received the Best Demo Award (Honorable Mention) at IEEE ISMAR 2026 in Bari, Italy, for "Flow Matching for Augmented Reality Applications," a real-time prompt-based AR stream editing demo that renders at 108…
A developer working on Launchpad by Axonscape published a fix for public forms that get flooded by spam bots, recommending Cloudflare Turnstile as an invisible check with server-side token verification using TURNSTILE_SE…
Baidu released ERNIE 5.5 Flash Preview, a foundation model with 280 billion total parameters and 14 billion activated per token, a 1 million token context window, and generation speeds above 200 tokens per second. Baidu …
Meta AI released Muse Ember, a model it says matches top models such as GPT-6 Luna and Claude Haiku at $0.10 per million input tokens and $0.50 per million output tokens for shorter prompts, placing it on the Artificial …
A new arXiv paper (2611.13441v1) measures a "tokenization premium" across 31 languages and 7 tokenizers released in 2023-2024, finding premiums reach 10x for Amharic in older tokenizers and remain at 6.8x in the newest c…
A research team reported in arXiv paper 2611.12933v1 that coding-agent infrastructure ran autonomous synchrotron X-ray experiments on nanoparticle superlattices across three experiments, with 61-67% of messages involving…
A research paper posted to arXiv (2611.13509v1) introduces Mobile AI Factory, an agentic system in which an orchestrator LLM builds a fine-tuned, quantized small language model for a specified task and target phone by ch…
A new arXiv paper (2611.13107v1) proposes online, deterministic verification of LLM reasoning steps, using code-based checkers such as a computer algebra system instead of a verifier LLM, so each step is validated in rea…
A new arXiv paper (2611.12995v1) proposes an ℓ1 regularization method that prunes unnecessary components from the target hidden states used in large language model knowledge editing, reducing interference with existing k…
Researchers posted arXiv:2611.12936v1, introducing LandslideFM, which they describe as the first unified foundation model connecting visual landslide perception tasks with contextual interpretation. LandslideFM uses a So…
Script Master Labs launched The Apprentice, a 30-day autonomous revenue experiment in which an AI agent sells AI-readiness audits for $0.15 USDC per call over the x402 protocol on Base, seeded with $100 USDC and capped b…
An open-source project called Restless, presented as a Show HN submission, runs a simulated company inside an AI harness in which five agents — including a product strategist and a claims auditor — plan and execute work …
A non-developer building a Mac app with AI coding tools like Claude to turn an iMac into a secondary display fixed two bugs reported by a beta tester, despite not owning the Intel iMac hardware involved. The fixes added …
Alibaba's Qwen3.5 Plus model is being deprecated under Alibaba Model Studio's deprecation policy, with a retirement date of 2027-01-19, according to the provider statement. The model, listed as qwen3.5-plus, supports tex…
An Australia-China academic collaboration published TongueReenact in July, a framework it describes as the first for cross-identity tongue dynamics transfer in face reenactment, combining 3D Gaussian Splatting, the FLAME…
Granola launched Agent Brief on November 11, a feature that generates a nightly-rewritten snapshot of a user's priorities, projects, contacts, glossary, communication style, habits and role from their meetings, which use…
Aegis Solis released a research dataset of 112 registered questions on AI coexistence, organized into families such as "Epistemic limits and self-knowledge" (code EPI) and tracked with fields including question_id, answe…
Hackers exploited multiple vulnerabilities in Railway's infrastructure, probing several entry points at once and analyzing the platform's internal source code to find new routes in, and generated $740,000 in fraudulent c…
An open-source privacy tool called Cloakroom was released to mask secrets and sensitive data from AI agents by swapping real values for placeholders and restoring the originals locally. The developer built Cloakroom afte…