Homepage
Close
Menu

Site Navigation

  • Home
  • Archive(TODO)
    • By Day
    • By Month
  • About(TODO)
  • Stats
Close
by OpenAI on 2019-03-06 (UTC).

In collaboration with Google, we're releasing Activation Atlases: a new technique for visualizing what interactions between neurons can represent.

đź’»Blog: https://t.co/nLfd0vpED4
📝Paper: https://t.co/qh2gozddHE
🔤Code: https://t.co/VwLN5j3uZo
🗺️Demo: https://t.co/9d1PmPqt6H pic.twitter.com/7c6Q6GvNLq

— OpenAI (@OpenAI) March 6, 2019
researchdataviz
by ch402 on 2019-03-06 (UTC).

For the last few years, one of the staples of my research has been visualizing individual neurons in vision models. But that's only a partial picture -- neurons work together.

Activation Atlases are a way to explore the space neurons jointly represent.https://t.co/TEa9WuYBMg pic.twitter.com/yb3zvpxmhq

— Chris Olah (@ch402) March 6, 2019
researchtooldataviz
by ch402 on 2019-03-06 (UTC).

The narrative around neural nets sometimes polarizes in two directions: they’re superhuman perfect systems, or cheap fragile hacks that break for stupid reasons.

I’d just like to say neural nets have issues, but gosh, there’s some really amazing stuff going on inside. pic.twitter.com/GzLrxE5bVA

— Chris Olah (@ch402) March 6, 2019
research
by JanelleCShane on 2019-03-06 (UTC).

Here's how they knew a baseball might fool their shark detector:
"In this example we see another detector that seems to be playing two roles: detecting red stitching on a baseball and a sharks’s white teeth and pink inner mouth." pic.twitter.com/skov3wAkTt

— Janelle Shane (@JanelleCShane) March 6, 2019
researchcv

Tags

learning tutorial misc nlp rstats gan ethics research dataviz survey python tool security kaggle video thought bayesian humour tensorflow w_code bias dataset pytorch cv tip application javascript forecast swift golang rl jax julia gnn causal surey diffusion
© Copyright Philosophy 2018 Site Template by Colorlib