ResearchAlpha Leak

Conferences Topics Top Authors Rankings Browse All

Home/Authors/Preetum Nakkiran

Preetum Nakkiran

Topic trends: 32,543 papers · similarity ≥ 0.4 · year ≥ 2024 · Data sourced from Semantic Scholar

34,598 papers | Abstracts: 31,650 (91.5%) | Citations: 34,598 (100.0%) | arXiv: 26,074 (75.4%)

Built: Feb 15, 2026, 9:33 AM AMS

10

papers

509

total citations

papers (10)

What Algorithms can Transformers Learn? A Study in Length Generalization

Revisiting Model Stitching to Compare Neural Representations

NEURIPS 2021arXiv

Normalizing Flows are Capable Generative Models

Smooth ECE: Principled Reliability Diagrams via Kernel Smoothing

A Formal Framework for Understanding Length Generalization in Transformers

LiDAR: Sensing Linear Probing Performance in Joint Embedding SSL Architectures

Knowledge Distillation: Bad Models Can Be Good Role Models

NEURIPS 2022arXiv

What You See is What You Get: Principled Deep Learning via Distributional Generalization

NEURIPS 2022arXiv

When Does Optimizing a Proper Loss Yield Calibration?

Benign, Tempered, or Catastrophic: Toward a Refined Taxonomy of Overfitting