Generalist agents — Gato and DeepMind's approach.
hardAnswer
- Reed et al. 2022 (Gato): single transformer trained on 600+ tasks (Atari, robotics, chat, image captioning) as tokenized sequences.
- Uses cross-attention over task tokens.
- Modest per-task performance, but shows one architecture can handle many domains.
- Followed by RT-2 (robotics VLA), OpenVLA, PI-0 (Pi Zero) — multi-task foundation models for embodied AI.
- Modern trend: transformer + massive data across modalities.
Check yourself — multiple choice
- Random
- Gato: single transformer trained on 600+ tasks as tokenized sequences (Atari, robotics, chat, image); RT-2 / OpenVLA / Pi-0 extend to embodied AI multi-task foundation models
- Same as GPT
- Not real
Generalist agents: Gato / RT-2 / OpenVLA / Pi-0; transformer across modalities.
#theory#applications
Practise Reinforcement Learning
214 interview questions in this topic.