M Edmonds

🛠 autonomous robotaxis @Cruise
🎓 UCLA CS PhD

[M Edmonds @mjedmonds_]

@TobinSouth @sarahookr No RIP m8

[M Edmonds @mjedmonds_]

@TobinSouth @divyasiddarth 🥲

[M Edmonds @mjedmonds_]

@jake__rl I have a solution for you

[M Edmonds @mjedmonds_]

@jake__rl I thought this platform was 💀

[M Edmonds @mjedmonds_]

@jake__rl Code = bugs

[M Edmonds @mjedmonds_]

@jake__rl what was the joke

[M Edmonds @mjedmonds_]

@fletchrichman @StonkyTonkPod I’m loving this discussion of multiplication tables

[M Edmonds @mjedmonds_]

@jake__rl Jake, I see you’ve tweeted

[M Edmonds @mjedmonds_]

Check our latest paper in Science Robotics! Robots that learn and understand our intention and values through interaction are critical to avoid many of the dystopian science fiction stories depicted in novels and the big screen :) https://t.co/jPhcAcbnJt

[M Edmonds @mjedmonds_]

@markjonesjr @russelljkaplan Interesting article…which part disagrees with this thread?

[M Edmonds @mjedmonds_]

@jake__rl That’s why I don’t write unit tests. More code = more bugs

[M Edmonds @mjedmonds_]

@jake__rl Important information it is

[M Edmonds @mjedmonds_]

@elonmusk @karpathy The paper clip optimizer problem

[M Edmonds @mjedmonds_]

@MIT_CSAIL Hierarchical, non-linear function approximation.

Only need 4 :)

[M Edmonds @mjedmonds_]

@PaulJBlazek @gradientpub Read the article, haven't dug into the paper yet, but at a high-level, what's the distinction between this work and neuro-symbolic methods?

[M Edmonds @mjedmonds_]

@donof_ Wild

[M Edmonds @mjedmonds_]

@jake__rl This level of tweet quality is why I don’t look at Twitter much. Negative reward

[M Edmonds @mjedmonds_]

@jake__rl @Twitter You can already do this with Lists