Shvetha

PhD Researcher | ML Engineer | ML for Science. I see AI as a handy toolbox! The better we understand these tools, the stronger our solutions will be.

[NASA Webb Telescope @NASAWebb]

Within this star-forming nebula (captured by Webb) are the smallest brown dwarfs ever detected. Neither planet nor star, these objects are formed like stars, raising questions about the star formation process. https://t.co/s58sxdx1dl

[Hubble @NASAHubble]

Group shot!

In this image of galaxy NGC 4258, @chandraxray data (royal blue) show superheated shockwaves created by black hole jets, along with Hubble's optical data (red, yellow, pale blue) and @NASAWebb's infrared view of dust filaments (orange): https://t.co/BHOg1GR5Qx

[Shvetha @shvetha_sc]

Super excited!

[NASA @NASA]

Before the James Webb Space Telescope can #UnfoldTheUniverse, it has to literally unfold. Find mission updates at:

- @NASAWebb (play-by-play)
- https://t.co/sGBPzkchmf (milestones)
- @NASA (livestreams)
- https://t.co/ht9HRe1k1U (major news)

Timeline: https://t.co/aY8ybuWvnW

[Shreya Shankar @sh_reya]

my college friend who also just started her PhD said "some people leave their jobs to go backpacking around the world for a few years. that's basically us except we're doing mental backpacking" and i can't stop thinking about how true that is

[Ferenc Huszár @fhuszar]

in my experience merge is a bit more like this

[Jia-Bin Huang @jbhuang0604]

How to make steady progress in my research?

I worked so damn hard but "IT JUST DOESN'T WORK!"😤
How can I unblock myself quickly and make good progress toward the goals?

Below I compiled a list of tips that I found useful. 👇

[hardmaru @hardmaru]

Signal’s ads on Instagram that revealed how specifically the ad itself was targeted, were banned by Facebook.

“It was evidently not OK with Signal exposing the personal data that Facebook has for sale to promote its own more private alternative.”

From: https://t.co/sCDCjCV5pR

[Mark Riedl @mark_riedl]

First draft 6 page limit

[Sara Hooker @sarahookr]

Yesterday, I ended up in a debate where the position was "algorithmic bias is a data problem".

I thought this had already been well refuted within our research community but clearly not.

So, to say it yet again -- it is not just the data. The model matters.

1/n

[hardmaru @hardmaru]

Brain2Pix: Fully convolutional naturalistic video reconstruction from brain activity 🧠🖼️

https://t.co/iJv92CP9s5

[hardmaru @hardmaru]

The Black-Scholes model, revised:

[Chip Huyen @chipro]

I updated my list of ML tools:

- 84 new tools (total 284) + interactive graph
- overview of MLOps landscape 2020
- ML tooling startups that have raised money in 2020. More than half are outside the Bay Area. Growing hubs: Boston, NYC, Tel Aviv.

https://t.co/QTr7eJvP9l

[Chip Huyen @chipro]

Is your model fast?
No
Is it cheap?
No
Does it at least solve our problem?
No
…
But it’s StAtE oF tHe ArT

[Chip Huyen @chipro]

Pandas is great for most day-to-day data analysis, but it has many quirks that can cause mysterious bugs or performance issues.

Here is a list of pandas things I’ve learned that have made my life so much easier. As always, feedback is much appreciated!

https://t.co/V4z8oedUfl

[Chip Huyen @chipro]

When talking to people who haven’t deployed ML models, I keep hearing a lot of misperceptions about ML models in production. Here are a few of them.

(1/6)

[I Am Devloper @iamdevloper]

> lets keep staging as close to prod as we can

me:

[PHD Comics @PHDcomics]

Day Off

[Shvetha @shvetha_sc]

@hugobowne Something that can be consumed without being cleaned!

[Jeff Dean @JeffDean]

AI & responsibility is a serious issue.

This blog post summarizes many @TensorFlow tools to help YOU create more responsible AI systems, looking at things like fairness, bias, interpretability, privacy & security.

Tools: https://t.co/DiAyL0B7zY
Blog: https://t.co/6b9DKNmenA

[hardmaru @hardmaru]

Great way to share https://t.co/C4T7xWREtv papers with notes, scribbles and annotations directly in the pdf! 🖍️

Feel free to contribute your annotated papers: https://t.co/LXZrdmcfqY