David Sheard

@mrtony_elt and I wrote a thing

Geodesics in the induced metric on a torus. Based on code by Matthew Cherrie and Thomas Waters demonstrations.wolfram.com/GeodesicsAndCo… (CC BY-NC-SA) #Maths

Bojo: "That's as far as my thinking has developed on this issue." I might credit him having thought about these complex and nuanced issues for a total of 5mins in his life, yet still he decides to 'treat' us to his unfiltered opinions #TransWomenAreWomen bbc.co.uk/sport/61012030

Write to your MP to tell them everyone fleeing the invasion of Ukraine should be able to apply to come to the UK, not just those with British close family members. Russian govt's nationalism began this crisis, don't let UK govt's nationalism turn away people in the greatest need.

I am immeasurably proud of @_AdamJarv as today he has been awarded Changemaker of the Year by @stonewalluk, and the @Tesco LGBTQ+ Network he chairs has been named Network Group of the Year 🥳👬🌈stonewall.org.uk/people/changem… stonewall.org.uk/people/network…

@DrTrapezio @James_A_Grant @stevenvmiller Love it! I'll add it to our list of things to discuss

Good gay god there's an R package of Rupaul's Drag Race data... rdrr.io/cran/dragracer… @stevenvmiller It's now my mission to find a legitimate way to shoehorn this into my modelling course. 😂🏳️‍🌈

@arxiv_cs_LG Just posted my first paper on arXiv with Ben Aslan and @dplatt_maths where we apply ideas from geometric topology to machine learning problems. Plus I get to throw in a whole bunch of ideas about Coxeter groups 🤩

Group invariant machine learning by fundamental domain projections. Benjamin Aslan, Daniel Platt, and David Sheard arxiv.org/abs/2202.02164

Papers, February 4, 2022:

“We feel that removing this has the potential to create an environment where gender prejudice and transphobic language is justified under the guise of academic freedom.” Read the full insightful statement 👇🏽 UCL UMC has so much to learn from our students!

Full statement from UCL's LGBTQ Equality Steering Group in response to UCL'S withdrawal from the two Stonewall schemes:

Can you find the mate in one, even if you don't know whose turn it is? ♘♟ #chalkdust14 chalkdustmagazine.com/features/pawns…

🚨 Chalkdust issue 14 has just launched! 🚨 Read it now! #chalkdust14 chalkdustmagazine.com/read/issue-14/

In just 167½ hours' time, #chalkdust14 will be released! Find out how to get a copy at chalkdustmagazine.com/blog/chalkdust….

Best toilet graffiti

@Condor_Ferries the 1970's called, they want their gender stereotypes and depictions of the female body back from your safety video #ThisDoesntFloatMyBoat

Congratulations to the unstoppable @ladyphyll on winning the @NatWest_Help *Lifetime Achievement* award at the 2021 #BritishLGBTAwards!

I had a pretty spectacular seat with @_AdamJarv and @Tesco at the #BritishLGBTAwards on Friday. Some very moving awards handed out.

Last night I ordered an @Uber, the driver accepted my booking and turned up, but then saw my wheelchair and drove off. This type of discrimination is still incredibly common - I wasn’t even surprised. Uber has no way of reporting drivers for this so it goes unchecked.

Congratulations to LGBTQ+ at @Tesco on winning the @OVOEnergy 🚗 Network Group or ERG 🚗 category at the 2021 #BritishLGBTAwards!

Like for covers this is easy, my guess would be that if G acts non freely, Stab_G(f(x))<=Stab_G(x), but is there more to it?

Maths/ggt people: if G acts nicely on nice spaces X and Y, when can a map f:X/G->Y/G be lifted to a map f':X->Y? In my case G=finite acting by linear isometries on real vector spaces but I'm interested in the answer a bit more generally. Where can I find this written down?