[summerlight]

I wonder what's the status quo on the non-LLM side; are we able to manually decode sound patterns to recognize dolphin's communication to some degree? If that's the case, I guess this may have a chance.