Theorinet Retreat 2022

/ October 28, 2022/ symposium

TimeSpeakerAffiliationTitle of Presentation

Wednesday, September 28

1 p.m.Sohir MaskeyLMU Munich'Stability to Deformations in Manifold Neural Networks.
1:15 p.m.Darshan ThakerJohns Hopkins UniversityGeneralization Analysis of Message Passing Neural Networks on Large Random Graphs
1:30 p.m.Muthu ChidambaramDukeAdaptive Conformal Inference Under Distribution Shift
1:45 p.m.John CherianStanfordProbabilistically Robust Learning: Balancing Average- and Worst-case Performance
2 p.m.Chenwei WuDukeT-Cal: An optimal test for the calibration of predictive models
2:15 p.m.Coffee Break
3 p.m. Alex RobeyPenn StateThe Value of Out-of-Distribution Data
3:15 p.m.Alex WeiBerkeleyConvolutional Filtering and Neural Networks with Non Commutative Algebras
3:30 p.m.Donghwan LeeUniversity of PennsylvaniaReverse Engineering ℓp attacks: A block-sparse optimization approach with recovery guarantees
3:45 p.m.Liangzu PengJohns HopkinsCollaborative Linear Bandits with Adversarial Agents: Near-Optimal Regret Bounds
4 p.m.Simon ZhaiBerkeleyComputational Benefits of Intermediate Rewards for Goal-Reaching Policy Learning
4:15 p.m. Yixuan TanDukeDiffusion of Information on Networked Lattices by Gossip
4:30 p.m. Isaac GibbsStanfordDistribution-free Prediction Sets Adaptive to Unknown Covariate Shift

Friday, September 30

2:15 p.m.Hans ReissUniversity of PennsylvaniaLearning by Transference: Training Graph Neural Networks on Growing Graphs
2:30 p.m.Chen XuDukeSpace-Time Graph Neural Networks
2:45 p.m. Teresa HuangJohns Hopkins UniversityGraph Neural Networks Are More Powerful Than We Think
3 p.m.Coffee Break
3:30 p.m.Ziqing XuJohns Hopkins UniversityMore Than a Toy: Random Matrix Models Predict How Real-World Neural Representations Generalize
3:45 p.m.Juan CervinoUniversity of PennslyvaniaGlobal Linear and Local Superlinear Convergence of IRLS for Non-Smooth Robust Regression
4:07 p.m.Alejandro Parada-MayorgaUniversity of PennslyvaniaTowards Understanding the Data Dependency of Mixup-Style Training
4:20 p.m.Zhiyang WangUniversity of PennslyvaniaDissecting Hessian: Understanding Common Structure of Hessian in Neural Networks
4:35 p.m.Ashwin De SilvaJohns Hopkins UniversityInvertible Neural Networks for Graph Prediction
Samar HadouUniversity of Pennslyvania
Charilaos KanatsoulisUniversity of PennslyvaniaFrom Local to Global: Spectral-Inspired Graph Neural Networks
Share this Post