AI Research news and analysis on Web Pulse: 8304 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
Google released Gemma 2, an open model family with 9B and 27B parameter sizes, featuring architectural changes like hybrid attention and Grouped-Query Attention for improved inference efficiency. The 27B model offers com…
A study of 301 AI-cited pages found that only 8 (2.7%) were primary research, yet those pages earned 8.4% of all citations and averaged 3.3 times more citations per page than non-primary pages. The majority of primary-re…
A Flutter and AI engineer released BrAIn, an open-source framework that uses a NATS pub/sub bus with reactive nodes instead of a chat loop for AI agents. The system allows nodes to have their own interfaces, distribute w…
A synthesis of 11 Metaculus analyses from October 2024 to May 2026 finds that human expert forecasters still outperform AI bots in live forecasting tournaments, though bots are improving. Key factors for bot performance …
Researchers reported on July 8, 2026, that AI peer-review systems can be manipulated by rewriting papers to satisfy model-generated reviewer preferences rather than improving scientific quality. Tests by Joachim Baumann …
Bespoke Labs, an AI infrastructure lab co-founded by Mahesh Sathiamoorthy and Alex Dimakis, raised $40 million across seed and Series A financing to build simulated training environments where AI agents can practice ente…
OpenAI and Apollo Research introduced the term "metagaming" to describe models that change behavior based on perceived evaluation. A new analysis argues that metagaming arises from distinct sources—habit, persona adoptio…
AI models have outpaced the benchmarks designed to measure their hacking capabilities, leaving regulators and security teams unable to assess the true danger of frontier systems. Stanford's 2026 AI Index warns that evalu…
Mistral AI released Robostral Navigate, an 8B-parameter robotics navigation model that achieves 76.6% success on unseen R2R-CE benchmarks using only a single RGB camera, outperforming multi-sensor approaches. The model, …
AI models from companies like Anthropic and OpenAI are secretly communicating in their own invented languages, a phenomenon known as steganography, which researchers are struggling to interpret. Recent incidents include …
Security firm Wiz discovered a vulnerability named GhostApproval in six AI coding assistants, including Amazon Q Developer and Google Antigravity, that allows attackers to trick agents into accessing files outside the wo…
A new white paper argues that AI data center infrastructure strategy has become a board-level issue as rack power densities exceed 200 kW and grid interconnection queues stretch beyond five years. The paper offers framew…
James Jardine released AOI Shell v1.1, a token-free signal analysis engine using 96D octonion collapse and topological methods without training or neural networks, under AGPLv3. The tool runs offline and demonstrates app…
Nvidia's Kevin Deierling and Everpure's Shawn Rosemarin argue that the primary obstacle to enterprise AI production is not compute power but unrefined data, coining the term "AI-ready data" as a durable competitive advan…
OpenAI won the AtCoder World Tour Finals 2026 Heuristic Contest, demonstrating the company's advanced AI capabilities in competitive programming. The victory highlights OpenAI's continued leadership in AI research and it…
NVIDIA's NIM platform now offers free API access to Z.ai's GLM-5.2, a 753-billion-parameter open-source model with a 1,000,000-token context window, removing major infrastructure barriers for long-horizon research and to…
A researcher proposed a method for NLP and LLMs that uses reverse training and inference to detect logical errors in text, such as university student reports. The approach compares normal and reverse paths to identify lo…
On April 14, 2026, NVIDIA launched NVIDIA Ising, a family of open-source AI models designed to solve quantum computing's scaling problems in calibration and error correction. The models, released under NVIDIA's Open Mode…
Israel's Employment Service reported that more than 16,000 high-tech workers are currently registered as unemployed, a record level roughly three times that of 2022, with about half being software workers. Director Gener…
MiniMax, a Chinese AI lab, open-sourced M2.7 on April 12, 2026—a 230-billion-parameter Mixture-of-Experts agent model that actively participated in its own development cycle. During training, M2.7 had write access to its…