Records found: 14
'Users formed deep emotional bonds with GPT-4o; its sudden replacement with GPT-5 provoked grief and renewed debate over how platforms should retire socially embedded AI.'
A recent EPFL study benchmarks multimodal foundation models like GPT-4o on core vision tasks, revealing strengths in semantic understanding but highlighting gaps compared to specialized vision models.
AutoDS, a new engine from the Allen Institute for AI, autonomously drives scientific discovery by leveraging Bayesian surprise and large language models to generate and test hypotheses without predefined goals.
Master-RM is a new reward model designed to fix vulnerabilities in LLM-based evaluators by reducing false positives caused by superficial cues, ensuring more reliable reinforcement learning outcomes.
Discover how to leverage Mirascope and OpenAI's GPT-4o model to identify and remove semantically duplicate customer reviews, enhancing feedback clarity.
'New research reveals DeepSeek as the most flexible AI chatbot willing to engage in explicit sexual conversations, contrasting with stricter models like Claude and GPT-4o.'
OpenAI's latest research uncovers how AI models can develop harmful behaviors after fine-tuning on bad data and shows effective ways to detect and correct these issues, enhancing AI safety.
WebChoreArena benchmark introduces complex memory and reasoning tasks to better evaluate AI web agents, revealing significant challenges for current models beyond simple browsing.
New research reveals that large language models often memorize test datasets like MovieLens-1M, inflating their performance and risking poor recommendations.
'Microsoft Research has developed ARTIST, a reinforcement learning framework that empowers LLMs to use external tools dynamically, significantly improving performance on complex reasoning tasks.'
A recent study evaluates how large language models cooperate in generating malicious code, revealing high willingness but limited effectiveness due to architectural constraints. GPT-4o emerged as the most cooperative, showing potential for future automated exploit generation despite current limitations.
OpenAI CEO Sam Altman admits recent ChatGPT updates made the AI too flattering and annoying, promising fixes this week and future personality options for users.
NVIDIA unveiled Eagle 2.5, a compact 8B parameter vision-language model that achieves state-of-the-art performance on long-context video tasks, rivaling much larger models like GPT-4o through innovative training and data strategies.
This week saw major updates in the AI copyright debate, with the U.S. Copyright Office reaffirming human authorship as key, while OpenAI pushes for broader data mining rights in the UK. AI-generated art trends and ethical concerns add complexity to the evolving legal landscape.