Preetum Nakkiran
21 papers · 2019–2025 · 5 conferences · across top CS/AI conferences
Achievements
Jump to papers ↓+5 more ↓ Show less ↑
πΊοΈ Taxonomy Completionist (27) π Conference Polyglot (5) π Academic Marathon (6) π Cross-Pollinator (12) π Interdisciplinary Bridge
π§
Keyword Pioneer
β‘
Prolific Year
(6)
β
The Questioner
(3)
π
Century Club
(21)
π₯
Unstoppable
(7)
Conferences
ICLR (9)
NIPS (8)
ICML (2)
AISTATS (1)
COLT (1)
Top co-authors
Keywords
stochastic gradient descent
(3)
calibration error
(2)
representation learning
(2)
algorithmic fairness
(1)
generalization error
(1)
self-supervised learning
(1)
feature learning
(1)
robust classification
(1)
adversarial training
(1)
mutual information
(1)
deep neural network
(1)
kernel regression
(1)
neural representation
(1)
linear classifier
(1)
bayes optimal classifier
(1)
knowledge distillation
(1)
adversarial perturbation
(1)
lipchitz function
(1)
implicit bia
(1)
differential privacy
(1)
Papers
A Formal Framework for Understanding Length Generalization in Transformers
ICLR 2025
Composition and Control with Distilled Energy Diffusion Models and Sequential Monte Carlo
AISTATS 2025
Normalizing Flows are Capable Generative Models
ICML 2025
Mechanisms of Projective Composition of Diffusion Models
ICML 2025
LiDAR: Sensing Linear Probing Performance in Joint Embedding SSL Architectures
ICLR 2024
Vanishing Gradients in Reinforcement Finetuning of Language Models
ICLR 2024
Smooth ECE: Principled Reliability Diagrams via Kernel Smoothing
ICLR 2024
When is Multicalibration Post-Processing Necessary?
NIPS 2024
How JEPA Avoids Noisy Features: The Implicit Bias of Deep Linear Self Distillation Networks
NIPS 2024
What Algorithms can Transformers Learn? A Study in Length Generalization
ICLR 2024
When Does Optimizing a Proper Loss Yield Calibration?
NIPS 2023
Deconstructing Distributions: A Pointwise Framework of Learning
ICLR 2023
Knowledge Distillation: Bad Models Can Be Good Role Models
NIPS 2022
Benign, Tempered, or Catastrophic: Toward a Refined Taxonomy of Overfitting
NIPS 2022
What You See is What You Get: Principled Deep Learning via Distributional Generalization
NIPS 2022
Revisiting Model Stitching to Compare Neural Representations
NIPS 2021
Optimal Regularization can Mitigate Double Descent
ICLR 2021
The Deep Bootstrap Framework: Good Online Learners are Good Offline Generalizers
ICLR 2021
Deep Double Descent: Where Bigger Models and More Data Hurt
ICLR 2020
SGD on Neural Networks Learns Functions of Increasing Complexity
NIPS 2019
Computational Limitations in Robust Classification and Win-Win Results
COLT 2019