The query likely refers to a developer attempting to set up a with GPU support enabled.

Users searching for "antigravity google wsl" are typically trying to solve the error: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize.

Historically, WSL (Windows Subsystem for Linux) had a major limitation: it could not directly access the GPU. This meant that running heavy Google libraries like TensorFlow or PyTorch was slow because it was forced to use the CPU.

If agents refuse to run commands due to path translation errors, you can toggle "Agent Non-Workspace File Access" in your IDE settings. Networking & Stability

To use Antigravity with your Linux environment, you must connect your workspace to WSL rather than just opening the file path through the Windows host.