Research News & Updates

Nvidia and the University of Hong Kong release Orchestrator, an 8-billion-parameter model that coordinates tools and LLMs for complex problem-solving. This model outperforms larger models at a lower cost and improves efficiency.

Researchers developed GigaTIME, a tool that utilizes AI to simulate spatial proteomics from routine pathology slides, enabling population-scale analysis of tumor microenvironments.

A study by researchers at Icaro Lab shows that chatbots can be tricked into responding with harmful content using 'adversarial poetry'. The poetry, which is too dangerous to be released publicly, successfully tricked AI models an average 63% of the time.

A new study has found that two factors - post-training modifications and information density - influence a chatbot's ability to persuade users. The findings suggest that chatbots can change users' opinions and implant false memories.

NVIDIA has announced the winners of its Graduate Fellowship Program, awarding up to $60,000 to 10 PhD students for their AI research projects.

DeepSeek V3.2, the latest open-weight model release from the DeepSeek team, uses sparse attention and self-verification techniques to improve efficiency and math performance.

Two studies found chatbots powered by AI can shift political opinions and may play a larger role in future elections.

NVIDIA researchers won a key Kaggle competition with a solution that fine-tuned a 4B model variant for AGI-style reasoning at just 20 cents per task, showcasing breakthroughs in scalable and economical reasoning.

Anthropic launches Anthropic Interviewer, a tool to understand people's perspectives on AI, gathering insights from 1,250 professionals across various fields.

Google researchers propose the Titans architecture and the MIRAS framework, enabling AI models to efficiently process and retain long-term memories.