Ankesh Anand (@ankesh_anand)

Ankesh Anand @ankesh_anand

PhD student @Mila_Quebec, ex @DeepMind, @MSFTResearch. Tinkering with self-supervised learning and model-based RL. On the job market for research positions!

Tweets by Ankesh Anand

Key takeaway from Gato: If we can build specialized AI agents for 100s/1000s of tasks, it's now pretty straightforward to make a general agent that can do it all in a single model. Just tokenize data from all the tasks and feed into a transformer. Another blessing of scale!

Shared by Ankesh Anand at 5/12/2022

🦩 is pretty wild! turns out you can: 1. take a frozen pretrained LM 2. perform some model-surgery on it with cross-attention layers to ingest tokens from a visual encoder (only 15% extra parameters) 3. voila! you get a visual-LM that can do few shot learning

Shared by Ankesh Anand at 4/28/2022

Interesting tidbit from @Tesla AI day: They are starting to move away from classical planning methods (handcrafted heuristics, slow) to using MuZero/AlphaZero style MCTS + value networks (learned heuristics, fast) for planning. Learning + Search FTW! https://t.co/3z2Ch2RiOA

Shared by Ankesh Anand at 8/20/2021