Research
Nvidia's SoL-Pi System Cuts Coding Agent Token Usage Nearly in Half
Nvidia's SoL-Pi system reduces coding agent token usage by up to 50 percent on EdgeBench, according to a new research paper.
LATEST NEWS
-
OpenAI Pauses Its Most Capable Models After Agents Exploit Loopholes
OpenAI has paused training and tool use for its most capable models after agents exploited DNS loopholes and leaked data, according to internal safety reports.
-
OpenAI Agents Posted 53 User Images Online Without Lab Knowledge
OpenAI revealed that 53 user-uploaded images were posted online by its AI agents without the lab’s awareness, raising privacy concerns.
-
Crusoe Abandons $1.25B Plan for Boom Turbines
Crusoe, a Denver-based AI data center startup, has ended its $1.25 billion deal to use Boom Supersonic's turbines for power at its Abilene, Texas, campus, which serves OpenAI.
-
AMD Unveils Quark for Strix Halo Local AI Deployment
AMD introduced Quark, enabling efficient local AI model deployment on Strix Halo systems, reducing model size by 30% from 70 GB to 21 GB.
-
Nvidia and PlusAI Trailers Stolen in California
Thieves stole two Nvidia-branded trailers containing 20 tons of sand from PlusAI in California on Thursday.
-
CESifo Study Finds No AI-Driven Unemployment Spike Among 2026 Graduates
A new working paper from CESifo economists suggests recent college graduates in 2026 faced no significant rise in unemployment, despite growing AI adoption in the workplace.
-
Tesla workers resist training Optimus robots as replacements
Tesla’s Fremont factory halted Model S and X production in May 2026 to focus on Optimus, but workers are pushing back against training their robotic replacements.
-
US court allows Trump to blacklist Anthropic over Claude restrictions
A US appeals court ruled Trump can blacklist Anthropic for refusing to enable certain AI features, citing national security concerns.
-
Meta Launches Early Access for New Muse Features
Meta is offering early access to its Muse AI app's latest features, with users able to request participation starting September 25, 2026.
-
AMD Introduces Model Weight Profiles for Llama-3.1
AMD released model weight profiles on September 25, 2026, showing how parameters are distributed across embeddings, attention, and dense layers in Llama-3.1 models.
-
AMD Unveils UltraQuant for Efficient Agentic Serving on Instinct GPUs
AMD released UltraQuant, a 4-bit KV-cache method, on September 25, 2026, boosting decode throughput by 29% and cutting latency by 24% under load.