OpenAI Details GPT-Red: An Internal Automated Red-Teaming Model That Beat Human Red-Teamers 84% To 13% On Prompt Injection
July 16, 2026admin NUAI,Committee,News,Uncategorized(0)This week, OpenAI published details of GPT-Red, an internal-only automated red-teaming model. Its job is...
OpenAI called the Hugging Face attack unprecedented. But we’ve been here before.
July 27, 2026admin NUAI,Committee,News,Uncategorized(0)This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories...
OpenAI Adds Chrome Extension to Codex, Letting Its AI Agent Access LinkedIn, Salesforce, Gmail, and Internal Tools via Signed-In Sessions
May 9, 2026admin NUAI,Committee,News,Uncategorized(0)OpenAI has launched a Codex Chrome extension for Mac and PC to streamline browser-based workflows...
Open World Knowledge Aided Single-Cell Foundation Model with Robust Cross-Modal Cell-Language Pre-training
January 12, 2026admin NUAI,Committee,News,Uncategorized(0)arXiv:2601.05648v1 Announce Type: cross Abstract: Recent advancements in single-cell multi-omics, particularly RNA-seq, have provided profound...
ONLY: One-Layer Intervention Sufficiently Mitigates Hallucinations in Large Vision-Language Models
July 2, 2025admin NUAI,Committee,News,Uncategorized(0)arXiv:2507.00898v1 Announce Type: cross Abstract: Recent Large Vision-Language Models (LVLMs) have introduced a new paradigm...
Online-PVLM: Advancing Personalized VLMs with Online Concept Learning
November 26, 2025admin NUAI,Committee,News,Uncategorized(0)arXiv:2511.20056v1 Announce Type: new Abstract: Personalized Visual Language Models (VLMs) are gaining increasing attention for...
Online harassment is entering its AI era
March 5, 2026admin NUAI,Committee,News,Uncategorized(0)Scott Shambaugh didn’t think twice when he denied an AI agent’s request to contribute to...
One-Token Rollout: Guiding Supervised Fine-Tuning of LLMs with Policy Gradient
October 1, 2025admin NUAI,Committee,News,Uncategorized(0)arXiv:2509.26313v1 Announce Type: new Abstract: Supervised fine-tuning (SFT) is the predominant method for adapting large...
One Trigger Token Is Enough: A Defense Strategy for Balancing Safety and Usability in Large Language Models
August 5, 2025admin NUAI,Committee,News,Uncategorized(0)arXiv:2505.07167v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) have been extensively used across diverse...
