US EditionHuman-Like Object Concept Representations Emerge Naturally in Multimodal Large Language Models
- On June 9, 2025, Chinese researchers published a paper in Nature Machine Intelligence demonstrating that multimodal large language models can spontaneously form object concept systems closely resembling those in humans.
- The study combined behavioral experiments with neuroimaging to explore whether AI models can form conceptual systems similar to human cognition.
- Researchers found that multimodal LLMs extract 66 concept dimensions strongly correlated with neural activity and outperform unimodal models in matching human behavior.
- He Huiguang highlighted that human intelligence fundamentally involves forming complex concepts of natural objects, encompassing not just their physical traits but also their functions, emotions, and cultural significance, whereas current models primarily capture semantic information with limited sensory understanding.
- The results suggest integrating multimodal sensory input could improve AI’s conceptual understanding, advancing human-like cognition in artificial systems.
11 Articles
11 Articles
Multimodal LLMs can develop human-like object concept representations: study
A group of Chinese scientists confirmed that multimodal large language models (LLMs) can spontaneously develop human-like object concept representations, providing a new path for the cognitive science of artificial intelligence (AI) and a theoretical
Chinese scientists confirm AI capable of spontaneously forming human-level cognition
Can artificial intelligence (AI) recognize and understand things like human being? Chinese scientific teams, by analyzing behavioral experiments with neuroimaging, have for the first time confirmed that multimodal large language models (LLM) based on artificial intelligence technology can spontaneously form an object concept representation system highly similar to that of humans; to put simply, artificial intelligence can spontaneously develop h…
Human-like object concept representations emerge naturally in multimodal large language models
Understanding how humans conceptualize and categorize natural objects offers critical insights into perception and cognition. With the advent of large language models (LLMs), a key question arises: can these models develop human-like object representations from linguistic and multimodal data? Here we combined behavioural and neuroimaging analyses to explore the relationship between object concept representations in LLMs and human cognition. We c…
The Manila Times
PolyU-led research reveals that sensory and motor inputs help large language models represent complex concepts
PolyU-led research reveals that sensory and motor inputs help large language models represent complex concepts
VeBrain: A Unified Multimodal AI Framework for Visual Reasoning and Real-World Robotic Control
Bridging Perception and Action in Robotics Multimodal Large Language Models (MLLMs) hold promise for enabling machines, such as robotic arms and legged robots, to perceive their surroundings, interpret scenarios, and take meaningful actions. The integration of such intelligence into physical systems is advancing the field of robotics, pushing it toward autonomous machines that don’t just see and describe but also plan and move within their envir…
Does AI know what it is like to have a body and experience the natural world?
Can AI perform just as well as humans when asked about experiences relating to the senses of the natural world and the human body? New publication reveals answers.It could be interesting to imagine what AI could be like if it really can understand the human experience in the body.Artificial intelligence (AI) such as in large language models (LLMs) cannot really comprehend the concept of forms in nature and sensations of the body the way us human…
Coverage Details
Bias Distribution
- 50% of the sources lean Left
To view factuality data please Upgrade to Premium
