paper
referenced-only
2023
paper:edelman-pareto-frontiers-in-deep-feature-learnin-2023Pareto frontiers in deep feature learning: Data, compute, width, and luck
ByBenjamin Edelman·Surbhi Goel·Sham Kakade·Eran Malach·Cyril Zhang
Related work— refs + corpus + external arXiv
Cited / in-corpus / arXiv badges show which signals surfaced each row. Multi-source rows weighted higher.
- ≈ 77%
- Identifying Functionally Important Features with End-to-End Sparse Dictionary LearningJordan Taylor, Nicholas Goldowsky-Dill, Lee Sharkey Dan Braun2024≈ 74%
- Large and Deep Factor ModelsBoris Kuznetsov, Semyon Malamud, Teng Andrea Xu, Yuan Zhang Bryan Kelly2026≈ 74%
- The Deepfake Detective: Interpreting Neural Forensics Through Sparse Features and ManifoldsJared Junkin Subramanyam Sahoo2025≈ 74%
- The Birth of Knowledge: Emergent Features across Time, Space, and Scale in Large Language ModelsMicah Adler, Nir Shavit Shashata Sawmya2025≈ 74%
- The Weight Gram Matrix Captures Sequential Feature Linearization in Deep NetworksDaniel Beaglehole, Adityanarayanan Radhakrishnan, Donghun Lee Taehun Cha2026≈ 73%
- Improving robustness of jet tagging algorithms with adversarial training: exploring the loss surfaceAnnika Stein2026≈ 73%
- Deeper or Wider: A Perspective from Optimal Generalization Error with Sobolev LossYahong Yang and Juncai He2026≈ 73%
- ≈ 73%
- Deep Learning of Part-based Representation of Data Using Sparse Autoencoders with Nonnegativity ConstraintsJacek M. Zurada, Olfa Nasraoui Ehsan Hosseini-Asl2016≈ 73%
- Semantic Convergence: Investigating Shared Representations Across Scaled LLMsSanjana Rathore, Andrew Rufail, Adrian Simon, Daniel Zhang, Soham Dave, Cole Blondin, Kevin Zhu, Sean O'Brien Daniel Son2025≈ 73%
- A Pattern Language for Machine Learning TasksIan Fan, Tuomas Laakkonen, Neil John Ortega, Thomas Hoffmann, Vincent Wang-Mascianica Benjamin Rodatz2025≈ 73%
- Multi-objective optimization determines when, which and how to fuse deep networks: an application to predict COVID-19 outcomesValerio Guarrasi and Paolo Soda2026≈ 72%
- Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retentionin corpus2026≈ 72%
- ≈ 72%
- Data, Depth, and Design: Learning Reliable Models for Skin Lesion AnalysisMichel Fornaciali, Afonso Menegola, Julia Tavares, Fl\'avia Vasques Bittencourt, Lin Tzy Li, Sandra Avila Eduardo Valle2026≈ 72%
- The Platonic Representation Hypothesisin corpus2024≈ 71%
- Finding Alignments Between Interpretable Causal Variables and Distributed Neural Representationsin corpus2023≈ 71%
- Learning without neurons in physical systemsin corpus2022≈ 70%
- Cross-Architecture Model Diffing with Crosscoders: Unsupervised Discovery of Differences Between LLMsin corpus2026≈ 70%
- Open Problems in Mechanistic Interpretabilityin corpus2025≈ 70%
- Mechanistic Knobs in LLMs: Retrieving and Steering High-Order Semantic Features via Sparse Autoencodersin corpus2026≈ 70%
- ≈ 70%
- ≈ 70%
- Monitoring and Discovering Reward Hacking with Internal Representations during LLM Evaluationsin corpus2026≈ 70%
- ≈ 70%
- ≈ 69%
- Toy Models of Superpositionin corpus2022≈ 69%
Similar preprints — Semantic Scholar
Cited by (2)
- Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention
Larger models learn tasks smaller models never learn, and the cause is data-induced competition for neurons, not expressivity: small models spend capacity on frequent or simple tasks, and their update
- Arithmetic in the Wild: Llama uses Base-10 Addition to Reason About Cyclic Concepts
Llama-3.1-8B solves cyclic arithmetic (e.g., "what month is six months after August?") not by performing modular addition in the period of the cyclic concept (12 for months, 7 for days of the week) as