● Covered by 1 source · 1 reportLow impact1 neutral
Discovered Materials (YC P26) uses AI agents to find new materials with specific properties
Key points
- AI agents discover novel crystalline materials.
- Targets include thermal conductivity, dielectric constant, Young's, and shear modulus.
- Agents use web search, coding sandbox, and ML tools.
- Proposes BEOL-compatible synthesis recipes.
AI-Driven Material Discovery
Discovered Materials, a Y Combinator P26 company, is utilizing AI agents to identify new materials. These agents are tasked with finding dynamically stable, novel, and BEOL-compatible crystalline materials that satisfy predefined targets for thermal conductivity, static dielectric constant, Young's modulus, and shear modulus. Each proposed material must also include a BEOL temperature and process-compatible synthesis recipe that an expert would deem feasible to attempt.
Agent Tools and Capabilities
The AI models are equipped with several tools to accomplish their task, mimicking those available to a computational materials scientist. These tools include web search capabilities, a coding sandbox with Python and Bash, and access to relevant materials science packages like pymatgen, mp_api, and ASE. Additionally, the agents use machine learning-based tools to compute dynamic stability, lattice thermal conductivity, static dielectric constant, and compliance tensor.
Benchmarking and ML Integration
The models operate without a stopping condition until an error occurs or their token budget of 100 million tokens is exhausted. The AI Security Institute's open-source Inspect framework is used to benchmark their performance. For property computations, the system leverages machine learning interatomic potentials (MLIPs), specifically the universal point edge transformer (UPET) foundation machine learning model PET-MAD. Future work may incorporate direct density functional theory calculations or a hybrid approach.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
Reporting from
Discovered Materials, a Y Combinator P26 startup, is developing AI agents to discover novel, dynamically stable, and BEOL-compatible crystalline materials that meet specific thermal, dielectric, and mechanical property targets. These AI agents are equipped with web search, a coding sandbox, and machine learning tools to compute material properties and propose synthesis recipes. This approach aims to automate and accelerate the discovery of new materials for advanced technological applications.