T-STAR is a style transfer approach that uses AMR graphs for intermediate representations. The first of it's kind, T-STAR yields high content preservation with negligible accuracy loss. Drop by the @emnlpmeeting Google booth at 3:30pm today to hear @JangraAnubhav talk about it!
INNOVATION
-
NeurIPS 2022 Becomes Longest Machine Learning Conference
By
–
#NeurIPS2022, the longest machine learning conference ever pic.twitter.com/ku7PVdclw7
— Gautam Kamath (@thegautamkamath) 10 décembre 2022#NeurIPS2022, the longest machine learning conference ever
-
Essential innovations beyond scaling in modern machine learning
By
–
Anyone who claims that the successes of modern machine learning are just due to scaling up needs to explain away each of these essential pieces:
– Adam
– Batch norm
– Resnets
– Contextual embeddings
– Transformers -
Backpropagation: Simple Optimization from Psychological Perspective
By
–
It's remarkable both that backprop is just a simple optimization scheme and that it took a psychologist to come up with it.
-
Google LaMDA Integration Threat to AI Writing Companies
By
–
can't wait to see how companies like Jasper and http://
copy.ai react when Google eventually integrates LaMDA into google docs if I'm at one of those places, the only thing I would be working on is how to differentiate before that day comes -
CICERO Learns Optimal Moves Using Reinforcement Learning
By
–
Meta AI’s @anton_bakhtin discusses how he and his team used Reinforcement Learning to teach CICERO the best moves for the millions of possible actions that could be made on any given board state. pic.twitter.com/W8SJKDz6Vf
— AI at Meta (@AIatMeta) 10 décembre 2022Meta AI’s @anton_bakhtin discusses how he and his team used Reinforcement Learning to teach CICERO the best moves for the millions of possible actions that could be made on any given board state.
-
Synthetic Data Innovation for Robust Financial Models
By
–
New ways of generating synthetic data to develop robust financial models is just one way that Mohit Jain at @BNYMellonWealth is driving data science innovation. Learn tips from him and 9 others in our new ebook! https://
domino.buzz/3gZiVwn -
Meta AI Reveals CICERO Components: Game Planning, Dialogue, Filtering
By
–
Meta AI’s @ml_perception reveals what was used to build CICERO: Theoretical game planning component Dialogue model Component filtering out any dialogue mistakes before messages are sent to players
-
TMLR explores pathway to major machine learning conferences
By
–
No concrete plans, but I'd like for there to be a path from TMLR to ML conferences. I suspect that, for TMLR, we'll experiment with this for smaller conferences first.
-
Crossing the Developer Chasm: Rethinking Devtools Adoption Strategy
By
–
Just recorded my theory of "Crossing the Developer Chasm" with @jacksbridger for the @ScalingDevTools podcast (feed https://
podcast.bitreach.io) "Early majority" vs "Late majority" terminology is 60 years old and isn't helpful for devtools startups. My take (tbc on @DXTipsHQ
):