Ben London

Machine learning researcher; scientist at Amazon Music. Don’t tweet much.

[derek guy @dieworkwear]

the most important thing about 1993 is that it was an amazing year for music

[Dan Roy @roydanroy]

What this program needs more of, IMO, is people committed to empirically grounded theory.

[Olivier Jeunen @olivierjeunen]

Offline RecSys evaluation is notoriously hard to get right. Unobserved confounders are often overlooked, but make it even harder.

Check out the teaser for my @CONSEQUENCES_ws talk at #RecSys2023 w/ @BenjoLondon!
@ACMRecSys @sharechatapp @AmazonScience
https://t.co/GwY5mzW1Vz

[Ben London @BenjoLondon]

@roydanroy Nina Mishra (not sure if she’s on Twitter)

[Ben London @BenjoLondon]

@haldaume3 Similar thoughts re: LLMs for recommendation. For instance, they can generate reasonable looking playlists based on a generic theme (eg “alt rock”)—which is incredible!—but may struggle with niche queries, and the playlists are usually kind of boring.

[Ben London @BenjoLondon]

Damn iPhone keyboard
Makes all my text messages
Come out as haikus

[Ben London @BenjoLondon]

GO BIRDS! 🦅

[Ben London @BenjoLondon]

@HudsonKate You’re pudding me on

[Ben London @BenjoLondon]

Thank goodness iOS can edit sent messages! Now I can stop texting about all these ducks and shot.

[Jonathan Stray @jonathanstray]

Open AI chatbot prompt: "Rewrite Baby Got Back in the Style of the Canterbury Tales."

Once upon a time in the land of Canterbury
A lady fair with a behind so hearty
Walked in the room with a waist so small
And a round thing in front of us all
...

[Ben London @BenjoLondon]

@tommy_da_cat They’re all over on the “Furry” server

[Ben London @BenjoLondon]

@liambaranauskas Bet his mouth is full of joyful turds tonight!

[Ben London @BenjoLondon]

@DimitrisPapail Would an example f be SGD? Like, you’re trying to learn the SGD update?

[Ben London @BenjoLondon]

@roydanroy @DimitrisPapail Or Rademacher complexity. With that, if each f_t is Lipschitz, you can recursively apply the contraction lemma, for which there’s a vector-valued version. But all the Lipschitz coeffs would multiply, so you’d hope they’re <1 so you don’t get O(C^T).

[Ben London @BenjoLondon]

Today at 12:05 @CONSEQUENCES_ws

[Ben London @BenjoLondon]

Fri, 9/23 @ 12:05 PDT: @AlexanderBuchh3 will present our work on a new estimator for offline evaluation of ranking policies.

https://t.co/VUj3AXcCud

[Ben London @BenjoLondon]

@olivierjeunen And now I know the workshop handle :)

[Ben London @BenjoLondon]

Collaboration with Giuseppe Benedetto & Thorsten Joachim, who might not be on Twitter

[Ben London @BenjoLondon]

Presenting 2 papers on new work at CONSEQUENCES+REVEAL Workshop @ RecSys ‘22

https://t.co/oVBB7GwVmU

Details 👇

[Ben London @BenjoLondon]

Fri, 9/23 @ 12:05 PDT: @AlexanderBuchh3 will present our work on a new estimator for offline evaluation of ranking policies.

https://t.co/VUj3AXcCud