Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Behnam Neyshabur Profile picture
Behnam Neyshabur
@bneyshabur
Research @AnthropicAI 💼 Past: Gemini @GoogleDeepMind (Co-led Blueshift team) 🧠 LLM Reasoning / AI Scientist 🎒Traveling & Backpacking -- All views are my own!
2 subscribers
Feb 5 • 6 tweets • 3 min read
I've left Anthropic to start something new.
🧵 Image Working at Anthropic was a wonderful experience.

Extremely high talent density, amazing culture, mission-driven, zero politics, leadership with real technical depth. Over the past year, I’ve learned so much about what made Anthropic successful and developed great respect for the founders and the team. I'm grateful to have been part of such an extraordinary organization.

Reflecting on the last 20 years, I see three phases—and I'm now entering a fourth:
Dec 21, 2022 • 10 tweets • 3 min read
These days, many people are interested in getting a PhD in ML. I think you should think really hard before committing to a PhD program in ML. Why?

I'm going to summarize some thoughts in this thread:

1/10 Graduate degree in ML is overrated. So is having publications in top ML venues. One can accomplish a lot in this field without any of these. The truth is that you don’t need to cover a lot of background before you can do interesting things in ML.

2/10
Jun 18, 2021 • 11 tweets • 4 min read
🆕 📰: Deep Learning Through the Lens of Example Difficulty

We introduce a measure of computational difficulty and show its surprising relationships with different deep learning phenomena.

Paper: arxiv.org/abs/2106.09647

with @Robert_Baldock & Hartmut Maennel

1/ Image ✅ We introduce a measure of computational example difficulty: the prediction depth (PD). PD is the earliest layer after which the network’s final prediction is already determined.

✅ We use k-NN classifier probes to determine the prediction of each layer (left panel).

2/ Image
Apr 30, 2021 • 7 tweets • 10 min read
Come to our talks and posters at #ICLR2021 to discuss our findings on understanding and improving deep learning! Talks and posters are available now! Links to the talks, posters, papers and codes in the thread:

1/7 When Do Curricula Work? (Oral at #ICLR2021)
with @XiaoxiaWShirley and @ethansdyer

Paper: openreview.net/forum?id=tW4QE…
Code: github.com/google-researc…
Video and Poster: iclr.cc/virtual/2021/p…

2/7
Jan 13, 2021 • 17 tweets • 4 min read
Some people say that one shouldn't care about publication and the quality matters. However, the job market punishes those who don’t have publications in top ML venues. I empathize with students and newcomers to ML whose good papers are not getting accepted. #ICLR2021
1/ Long thread at the risk of being judged:

I just realized that in the last 6 years, 21 of my 24 papers have been accepted to top ML conf in their FIRST submission even though the majority of them were hastily-written borderline papers (not proud of this). How is this possible?
2/