
Wow. 1 in 3 hospitalized with #COVID19 gets readmitted to the hospital within a few months after discharge. ~48,000 UK patients.
Great animated lecture on #LongCovid by @Dr2NisreenAlwan, animated by @VickiGSP using info from UK @IndependentSage experts.


Lecture again by @Dr2NisreenAlwan

LONG COVID\u201430% of hospital recovered #COVID19 patients end up back in hospital in <5 months; up to 12% die of complications. \u201cwe really need to prepare for #LongCovid. It\u2019s a mammoth task to follow up w/ these patients, but monitoring needs to be arranged\u201dhttps://t.co/h0y8WUn8sQ pic.twitter.com/Pk8GhQc9J5
— Eric Feigl-Ding (@DrEricDing) January 18, 2021
Psssst... pass it on:
— Eric Feigl-Ding (@DrEricDing) January 17, 2021
\u27a1\ufe0f COVID LUNG is worse than smoker\u2019s lung.
"Post-COVID lungs look worse than any type of terrible smoker's lung we've ever seen. And they collapse. And they clot off. And the shortness of breath lingers\u201d says Dr @BKendallMD. #COVID19 pic.twitter.com/h3PrzzN3rR
More from Eric Feigl-Ding

2) Here is what is really going to happen... most countries are having a gentle case decline with R(e) currently around 0.9. But this is deceiving. The #B117 is still relatively rare so far, so the R is being influenced mostly by the old common variant. But not for long...

3) Here is what is going to happen... currently R is ~0.9 in many places, but with the more infectious #B117, the R will jump 50% approximately. And it is inevitable (all CDC and Danish models say this) that B117 will take over as the reigning dominant variant soon...

4) and when that happens, what worked before to keep the pandemic contained at R of 0.9 will no longer work. Here is the model for Alberta, 🇨🇦 by @GosiaGasperoPhD. The B117 dotted red line will soon dominate and drive a new surge in latter half of March and April.

5) And Denmark 🇩🇰 CDC has found the same thing. I GQR works now for keeping R around 0.9 or even 0.8, will absolutely not work anymore once #B117 variant takes over. Forget about it. We will be hit hard. But there is a way—if we suppress R to 0.7 or less.
https://t.co/gOq0put4H5

2) That said, the new UK mutated strain is a more transmissible (contagious) variant. Keep transmissibility and illness severity separate. They are two different things. Don’t get them confused.
3) The new UK variant severity study is included in today’s WHO report, to be released soon this afternoon according to WHO sources. Waiting for it to drop. I’ll keep folks posted on the details.
4) Update: here is UK Technical report on #SARSCoV2 variant of concern (B.1.1.7) in 🇬🇧 with prelim findings from their case-control study of no increased severity for the new variant that I reported yesterday.
5) Also keep in a mind that a virus that spreads faster (more contagious) yields much worse total outcomes than a virus that is just merely more severe.
All things equal, a new mutated variant that is more contagious & no more severe is worse. And that is what we now have.
Why a SARS-CoV-2 variant that's 50% more transmissible would in general be a much bigger problem than a variant that's 50% more deadly. A short thread... 1/
— Adam Kucharski (@AdamJKucharski) December 28, 2020
HHS @SecBecerra has tested positive for COVID-19 again, the agency just announced. Last time he tested positive was May 18, after attending the G-7 Summit.
— Ariel Cohen (@ArielCohen37) June 13, 2022
2) #BA5 and #BA4 are worrisome. They are surging and they have high reinfection potential. Your past BA1/BA2 doesn’t substantially protect you from #Ba5/4
\U0001f4c8Surging #BA5 & #BA4 variants in both US & England. It\u2019s looking like #Ba5 is the new fastest strain\u2014which has ~10x increased in England in 1 month. Also notice that older #Omicron #B11529/#BA1 are already near *extinct*. I worry the \u201cnew\u201d Omicron BA1 bivalent vaccine is behind. pic.twitter.com/kLNH0gzPk3
— Eric Feigl-Ding (@DrEricDing) June 9, 2022
3) Excess deaths—the new #BA4 & #BA5 variants of the coronavirus are currently the **fastest growing** strains in the US & UK. ➡️They are exponentially replacing all other past strains. Learn from South Africa’s early warning signs and their excess deaths. #CovidIsNotOver #COVID
To anyone pointing to South Africa to suggest the impact of the omicron wave has been 'mild' or that the BA.4/5 haven't had much impact, I'd urge you to look at excess deaths. 29,500 excess deaths since Jan (omicron wave) & a peak of 1,844 excess deaths/wk during BA.4/5 wave.\U0001f9f5 pic.twitter.com/r7kLzmD5dG
— Dr. Deepti Gurdasani (@dgurdasani1) June 11, 2022
4) Is it a Paxlovid rebound for Becerra, as some are asking? Unclear—Paxlovid rebound usually only happens at day 10-14 that we know of and have data. The FDA says it’s 1-2% rebound but I counted 12% from this Pfizer graph submitted to FDA.
4) Look at these 97 Paxlovid patients here with viral RNA load data over time. Of 97, i conservatively counted at least 12 of 97 (12%) with viral load rebound after day 10. Even more if you count rebound after day 5. Thus, FDA\u2019s 1-2% doesn\u2019t match up here. https://t.co/bNoqj5W3Sk pic.twitter.com/B3Kh0Lc1A4
— Eric Feigl-Ding (@DrEricDing) May 5, 2022
5) and yes, #ba5 / #BA4 are very problematic. they are 2x more resistant than even BA2 for neutralizing the virus compared to those who had breakthrough reinfections from older Omicron.
Unfortunately neutralization experiments found that BA.4/5 is 2x more resistant than BA.2 to the blood obtained from vaccinated individuals who had a breakthrough infection with BA.1 or BA.2. 4/ pic.twitter.com/MEOPXuqTgg
— Dr. Jeff Gilchrist (@jeffgilchrist) June 13, 2022
More from All
Introducing Voltron: Language-Driven Representation Learning for Robotics!
Paper: https://t.co/gIsRPtSjKz
Models: https://t.co/NOB3cpATYG
Evaluation: https://t.co/aOzQu95J8z
🧵👇(1 / 12)

Videos of humans performing everyday tasks (Something-Something-v2, Ego4D) offer a rich and diverse resource for learning representations for robotic manipulation.
Yet, an underused part of these datasets are the rich, natural language annotations accompanying each video. (2/12)
The Voltron framework offers a simple way to use language supervision to shape representation learning, building off of prior work in representations for robotics like MVP (https://t.co/Pb0mk9hb4i) and R3M (https://t.co/o2Fkc3fP0e).
The secret is *balance* (3/12)
Starting with a masked autoencoder over frames from these video clips, make a choice:
1) Condition on language and improve our ability to reconstruct the scene.
2) Generate language given the visual representation and improve our ability to describe what's happening. (4/12)
By trading off *conditioning* and *generation* we show that we can learn 1) better representations than prior methods, and 2) explicitly shape the balance of low and high-level features captured.
Why is the ability to shape this balance important? (5/12)
You May Also Like
Covering one of the most unique set ups: Extended moves & Reversal plays
Time for a 🧵 to learn the above from @iManasArora
What qualifies for an extended move?
30-40% move in just 5-6 days is one example of extended move
How Manas used this info to book
The stock exploded & went up as much as 63% from my price.
— Manas Arora (@iManasArora) June 22, 2020
Closed my position entirely today!#BroTip pic.twitter.com/CRbQh3kvMM
Post that the plight of the
What an extended (away from averages) move looks like!!
— Manas Arora (@iManasArora) June 24, 2020
If you don't learn to sell into strength, be ready to give away the majority of your gains.#GLENMARK pic.twitter.com/5DsRTUaGO2
Example 2: Booking profits when the stock is extended from 10WMA
10WMA =
#HIKAL
— Manas Arora (@iManasArora) July 2, 2021
Closed remaining at 560
Reason: It is 40+% from 10wma. Super extended
Total revenue: 11R * 0.25 (size) = 2.75% on portfolio
Trade closed pic.twitter.com/YDDvhz8swT
Another hack to identify extended move in a stock:
Too many green days!
Read
When you see 15 green weeks in a row, that's the end of the move. *Extended*
— Manas Arora (@iManasArora) August 26, 2019
Simple price action analysis.#Seamecltd https://t.co/gR9xzgeb9K