Bilisim PostasiBLOG
← HomeAI Agents & Automation

Artificial Intelligence Agents can now perform Transistor Design in 2D Materials

· Source

Artificial Intelligence Agents can now perform Transistor Design in 2D Materials
Photo: BoliviaInteligente / Unsplash

Artificial intelligence researchers exceed a new limit in transistor design. The domain transistors in two-dimensional materials (FET) are traditionally drawn by researchers, now language model agents can automate these complex tasks. Starting from the obtained microphotograph images, a new benchmark is offered to accurately position the contact transistor contacts and doors.

2D-FET-Bench V2 includes basic points:

  • 128 transistor design task is formed using two-dimensional layer contours derived from actual microphotograph data Tags
  • For the first time, the language model agents are measured to the specific flake of specific transistor construction
  • The technical design process, which is forced to match the location and shape of the layer in optical micrograph
  • The potential to switch to an automated method can reduce the time intensive labor of the researchers

This development shows that artificial intelligence agents can carry out physical design and engineering tasks beyond abstract spatial logical thought. Thanks to Benchmark testing, performance of different language model agents can be objectively comparable. Such criteria can open practical applications in material science and nanotechnology of future AI systems.

The study proves that artificial intelligence can only be applied in more advanced areas than text processing.

#Transistor Design #Language Models #2D Materials #Machine Learning