AI Daily – 2025-08-05(Morning)
MLLMs Lack Core Cognition, Industry’s First Benchmark Released : Latest research reveals that Multimodal Large Language Models (MLLMs) generally lag humans by 10-30% in 12 core cognitive abilities, in
AI Daily – 2025-08-04(Evening)
DeepMind CEO Demis Hassabis on the Future of AGI and Science: In a recent interview, DeepMind CEO Demis Hassabis delved into the future of AGI, stating that AI can efficiently model all natural patter
AI Daily – 2025-08-04(Morning)
China’s AI Model and Agent Capability Breakthroughs: Zhipu’s GLM-4.5 model has been released, adopting an MoE architecture to strengthen Agent capabilities; Alibaba Cloud’s Qwen3 Coder Flash 30B and Z
AI Daily – 2025-08-03(Evening)
AI’s Mathematical Proof Capabilities Breakthrough: IMO Gold Medal and Formal Verification : Tsinghua alumni Yang Lin and Huang Yichen, using only prompt engineering, successfully enabled Gemini 2.5 Pr
AI Daily – 2025-08-03(Morning)
DeepSeek’s Next-Gen Tech Revealed Early: Liang Wen-feng’s Paper Wins ACL 2025 Best Paper Award : The DeepSeek team won the Best Paper Award at ACL 2025 for their Native Sparse Attention (NSA) mechanis
AI Daily – 2025-08-02(Evening)
AGENTSAFE, the First Safety Benchmark for Embodied AI Agents, Released : Beihang University, Zhongguancun Laboratory, Nanyang Technological University, and other institutions jointly released AGENTSAF
AI Daily – 2025-08-02(Morning)
Anthropic Discovers LLM “Personality Vectors” and Proposes New Training Paradigm : Anthropic’s latest research reveals the existence of specific neural activity patterns in large language models assoc
AI Daily – 2025-08-01(Evening)
Gemini 2.5 Deep Think IMO Gold Medal Model Released: Google DeepMind has released the Gemini 2.5 Deep Think model, which achieved gold-medal level performance in the International Mathematical Olympia
AI Daily – 2025-08-01(Morning)
OpenAI’s Research Direction and GPT-5 Outlook: OpenAI Chief Scientist Jakub Pachocki and Head of Research Mark Chen revealed the company’s progress on GPT-5 development and their views on AGI in an ex