AI Agent Elements Claw Completes Superconducting Material Development
As artificial intelligence pushes the boundaries of scientific exploration, a significant milestone has been achieved. On July 3, Alibaba DAMO Academy, in partnership with Renmin University of China and the University of Chinese Academy of Sciences, unveiled Elements Claw, the world’s first AI agent dedicated to discovering superconducting materials. This breakthrough signifies a shift in scientific discovery, moving AI from a supporting role to an independent research partner, while establishing an efficient automated framework for new material development.
Traditional methods for finding superconductors are notoriously slow and rely heavily on trial and error. Although major international databases like SuperCon have accumulated decades of data, they contain only about 2,000 recorded materials. Elements Claw overcomes this limitation through a hybrid architecture that combines specialized and general capabilities. Trained on a database of 125 million molecules and crystal structures, it utilizes a 10-billion-parameter atomic foundation model (Elements). This allows the system to accurately assess superconducting potential, achieving an AUC score of 0.996 and keeping critical temperature prediction errors within 1K.

Elements Claw operates with the comprehensive workflow of human scientists. It independently accesses vast literature, evaluates synthesis feasibility, designs experimental protocols, and enables algorithmic self-evolution upon identifying new insights. In practical applications, the AI efficiently screened 68,000 superconducting candidates from 2.4 million crystal structures, completing the task in just 28 GPU hours.
Experimental validation has confirmed four new superconducting materials, including HfZrRe4, which was designed entirely by AI, as well as Hf21Re25, Zr4VRe7, and Zr3ScRe8, discovered through the correction and deep analysis of existing databases. These materials exhibit critical temperatures as high as 6.5K.
Rong Yu, head of the Science Intelligence department at DAMO Academy, noted that these results validate the substantial potential of AI agents in material discovery. To accelerate industry progress, the team has open-sourced data for 2.4 million stable crystals. Professor Huang Wenbing from Renmin University’s Gaoqiang Institute of Artificial Intelligence highlighted that this intelligent framework is expected to be widely applied to other critical materials, such as solid-state battery electrolytes, multiphase catalysts, and thermoelectric materials, marking a new era in scientific research.
Related article
How Unitree is Shaping the Future of Humanoid Robotics
Unitree’s New Creature of Embodied AI with ultra-wide 4D LiDAR technology for advanced real-world navigation. Credit: UnitreeWang Xingxing, CEO of Unitree, targets world model breakthroughs to help humanoid machines execute 80% of tasks when placed i
Why Abnormal AI Partnered With OpenAI on Daybreak
Michael Aiello, Head of Product for Cyber at OpenAI | Credit: Michael Aiello/LinkedInAbnormal AI joins OpenAI’s Daybreak programme, as Mike Aiello says the partnership will accelerate the adoption of practical defences for enterprisesLeading frontier
LFM2.5 DSpark Draft Model Released: Inference Speed Boosted 3.18x
Liquid AI and Hugging Face have officially released the DSpark draft model checkpoints for the LFM2.5 series, covering LFM2.5-1.2B-Instruct, LFM2.5-2.6B, and LFM2.5-8B-A1B. This innovation introduces a novel speculative decoding path that significant
Related Special Topic Recommendations
Comments (0)
0/500
As artificial intelligence pushes the boundaries of scientific exploration, a significant milestone has been achieved. On July 3, Alibaba DAMO Academy, in partnership with Renmin University of China and the University of Chinese Academy of Sciences, unveiled Elements Claw, the world’s first AI agent dedicated to discovering superconducting materials. This breakthrough signifies a shift in scientific discovery, moving AI from a supporting role to an independent research partner, while establishing an efficient automated framework for new material development.
Traditional methods for finding superconductors are notoriously slow and rely heavily on trial and error. Although major international databases like SuperCon have accumulated decades of data, they contain only about 2,000 recorded materials. Elements Claw overcomes this limitation through a hybrid architecture that combines specialized and general capabilities. Trained on a database of 125 million molecules and crystal structures, it utilizes a 10-billion-parameter atomic foundation model (Elements). This allows the system to accurately assess superconducting potential, achieving an AUC score of 0.996 and keeping critical temperature prediction errors within 1K.

Elements Claw operates with the comprehensive workflow of human scientists. It independently accesses vast literature, evaluates synthesis feasibility, designs experimental protocols, and enables algorithmic self-evolution upon identifying new insights. In practical applications, the AI efficiently screened 68,000 superconducting candidates from 2.4 million crystal structures, completing the task in just 28 GPU hours.
Experimental validation has confirmed four new superconducting materials, including HfZrRe4, which was designed entirely by AI, as well as Hf21Re25, Zr4VRe7, and Zr3ScRe8, discovered through the correction and deep analysis of existing databases. These materials exhibit critical temperatures as high as 6.5K.
Rong Yu, head of the Science Intelligence department at DAMO Academy, noted that these results validate the substantial potential of AI agents in material discovery. To accelerate industry progress, the team has open-sourced data for 2.4 million stable crystals. Professor Huang Wenbing from Renmin University’s Gaoqiang Institute of Artificial Intelligence highlighted that this intelligent framework is expected to be widely applied to other critical materials, such as solid-state battery electrolytes, multiphase catalysts, and thermoelectric materials, marking a new era in scientific research.
How Unitree is Shaping the Future of Humanoid Robotics
Unitree’s New Creature of Embodied AI with ultra-wide 4D LiDAR technology for advanced real-world navigation. Credit: UnitreeWang Xingxing, CEO of Unitree, targets world model breakthroughs to help humanoid machines execute 80% of tasks when placed i
Why Abnormal AI Partnered With OpenAI on Daybreak
Michael Aiello, Head of Product for Cyber at OpenAI | Credit: Michael Aiello/LinkedInAbnormal AI joins OpenAI’s Daybreak programme, as Mike Aiello says the partnership will accelerate the adoption of practical defences for enterprisesLeading frontier
LFM2.5 DSpark Draft Model Released: Inference Speed Boosted 3.18x
Liquid AI and Hugging Face have officially released the DSpark draft model checkpoints for the LFM2.5 series, covering LFM2.5-1.2B-Instruct, LFM2.5-2.6B, and LFM2.5-8B-A1B. This innovation introduces a novel speculative decoding path that significant





Home






