3 ms·
Yes, here's an example of the "cheese" strategy: https://www.youtube.com/watch?v=R85WVTFPmRE https://www.youtube.com/watch?v=R85WVTFPmRE Plays "normally" (caut
by adotbacon 5y ago
Yes, here's an example of the "cheese" strategy: https://www.youtube.com/watch?v=R85WVTFPmRE https://www.youtube.com/watch?v=R85WVTFPmRE
Plays "normally" (cautiously since the bot is so good) until 7:00 accumulating gold for enough items to more safely eliminate the creeps: https://www.youtube.com/watch?v=R85WVTFPmRE&t=420s https://www.youtube.com/watch?v=R85WVTFPmRE&t=420s
- sillysaurusx 5y agoThank you so much! I couldn’t find a video anywhere I looked, and started wondering whether I imagined the whole thing. I think these kinds of adversarial examples will be extremely common in production models. People won’t be crafting images that fool the model into thinking you’re a stop sign; they’ll discover that when the human isn’t paying attention, you can run in front of a Tesla with a group of friends and it veers into oncoming traffic. (Terrible made-up example, but I’m pretty sure that it’s a losing game to play “can we think of all possible cases we need to train for ahead of time?”)
- hatsunearu 5y agothat's such a weird behavior, the bot completely ignores the opponent when in that particular spot?!