Category: Large Language Models
-
OpenAI’s GPT-5.4 Thinking Model Achieves Human Expert Parity on GDPVal Benchmark
OpenAI’s GPT-5.4 ‘Thinking’ model has scored 83.0% on the GDPVal benchmark, marking the first time a public AI…
-
Benchmark Breakdown: How Multimodal AI Models Are Redefining Evaluation Standards
For years, the AI landscape was neatly compartmentalized. Language models were judged on text-based tasks like question answering…
-
Nintendo Switch 2 Update Adds Handheld Mode Boost for 1080p Backward Compatibility
A new system update for the Nintendo Switch 2 introduces Handheld Mode Boost, enabling original Switch games to…
-
Ars Technica Retracts AI Agent Hit Piece Story Following Editorial Review
Ars Technica has withdrawn a story about an AI agent publishing a targeted article after a code rejection,…
-
CISA Mandates Patching for Three iOS Flaws Exploited by Sophisticated Coruna Kit
Federal agencies must address critical iOS vulnerabilities exploited by the Coruna hacking kit, which assembled 23 exploits into…
-
Invisible Unicode Characters Enable Stealthy Supply-Chain Attacks on GitHub, NPM
Security researchers uncover 151 malicious packages using invisible Unicode code to evade detection in GitHub and other repositories,…
-
TeamPCP’s Supply Chain Attack Deploys Self-Propagating Worm via Compromised Trivy Versions
A new hacking group, TeamPCP, has launched a persistent campaign using a self-propagating backdoor and a data wiper…
-
Nvidia GPU Rowhammer Exploits Escalate to Full Host Machine Takeover
Two independent research teams have demonstrated novel Rowhammer attacks on Nvidia Ampere GPUs, enabling full root control of…