Enric Boix-Adserà

According to our database, Enric Boix-Adserà authored at least 32 papers between 2018 and 2025.

Collaborative distances:

• Dijkstra number of three.
• Erdős number of three.

2025

FACT: the Features At Convergence Theorem for neural networks.

Enric Boix-Adserà, Neil Mallinar, James B. Simon, Mikhail Belkin

CoRR, July, 2025

Let Me Think! A Long Chain-of-Thought Can Be Worth Exponentially Many Short Ones.

Parsa Mirtaheri, Ezra Edelman, Samy Jelassi, Eran Malach, Enric Boix-Adserà

CoRR, May, 2025

The power of fine-grained experts: Granularity boosts expressivity in Mixture of Experts.

Enric Boix-Adserà, Philippe Rigollet

CoRR, May, 2025

Aggregate and conquer: detecting and steering LLM concepts by combining nonlinear predictors over multiple layers.

Daniel Beaglehole, Adityanarayanan Radhakrishnan, Enric Boix-Adserà, Mikhail Belkin

CoRR, February, 2025

On the inductive bias of infinite-depth ResNets and the bottleneck rank.

Enric Boix-Adserà

CoRR, January, 2025

2024

Characterizations of how neural networks learn

Enric Boix-Adserà

PhD thesis, 2024

Towards a theory of model distillation.

Enric Boix-Adserà

CoRR, 2024

When can transformers reason with abstract symbols?

Enric Boix-Adserà, Omid Saremi, Emmanuel Abbe, Samy Bengio, Etai Littwin, Joshua M. Susskind

Proceedings of the Twelfth International Conference on Learning Representations, 2024

Prompts have evil twins.

Rimon Melamed, Lucas H. McCabe, Tanay Wakhare, Yejin Kim, H. Howie Huang, Enric Boix-Adserà

Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, 2024

2023

Polynomial-time algorithms for multimarginal optimal transport problems with structure.

Jason M. Altschuler, Enric Boix-Adserà

Math. Program., May, 2023

Tight conditions for when the NTK approximation is valid.

Enric Boix-Adserà, Etai Littwin

Trans. Mach. Learn. Res., 2023

PROPANE: Prompt design as an inverse problem.

Rimon Melamed, Lucas H. McCabe, Tanay Wakhare, Yejin Kim, H. Howie Huang, Enric Boix-Adserà

CoRR, 2023

The NTK approximation is valid for longer than you think.

Enric Boix-Adserà, Etai Littwin

CoRR, 2023

Transformers learn through gradual rank increase.

Emmanuel Abbe, Samy Bengio, Enric Boix-Adserà, Etai Littwin, Joshua M. Susskind

Proceedings of the Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023, 2023

SGD learning on neural networks: leap complexity and saddle-to-saddle dynamics.

Emmanuel Abbe, Enric Boix Adserà, Theodor Misiakiewicz

Proceedings of the Thirty Sixth Annual Conference on Learning Theory, 2023

2022

Wasserstein Barycenters Are NP-Hard to Compute.

Jason M. Altschuler, Enric Boix-Adserà

SIAM J. Math. Data Sci., 2022

GULP: a prediction-based metric between representations.

Enric Boix-Adserà, Hannah Lawrence, George Stepaniants, Philippe Rigollet

Proceedings of the Advances in Neural Information Processing Systems 35: Annual Conference on Neural Information Processing Systems 2022, 2022

On the non-universality of deep learning: quantifying the cost of symmetry.

Emmanuel Abbe, Enric Boix-Adserà

Proceedings of the Advances in Neural Information Processing Systems 35: Annual Conference on Neural Information Processing Systems 2022, 2022

The merged-staircase property: a necessary and nearly sufficient condition for SGD learning of sparse functions on two-layer neural networks.

Emmanuel Abbe, Enric Boix Adserà, Theodor Misiakiewicz

Proceedings of the Conference on Learning Theory, 2-5 July 2022, London, UK., 2022

2021

Wasserstein barycenters can be computed in polynomial time in fixed dimension.

Jason M. Altschuler, Enric Boix-Adserà

J. Mach. Learn. Res., 2021

Hardness results for Multimarginal Optimal Transport problems.

Jason M. Altschuler, Enric Boix-Adserà

Discret. Optim., 2021

The staircase property: How hierarchical structure can guide deep learning.

Emmanuel Abbe, Enric Boix-Adserà, Matthew S. Brennan, Guy Bresler, Dheeraj Nagaraj

Proceedings of the Advances in Neural Information Processing Systems 34: Annual Conference on Neural Information Processing Systems 2021, 2021

Chow-Liu++: Optimal Prediction-Centric Learning of Tree Ising Models.

Enric Boix-Adserà, Guy Bresler, Frederic Koehler

Proceedings of the 62nd IEEE Annual Symposium on Foundations of Computer Science, 2021

2020

Graph Powering and Spectral Robustness.

Emmanuel Abbe, Enric Boix-Adserà, Peter Ralli, Colin Sandon

SIAM J. Math. Data Sci., 2020

High-precision Wasserstein barycenters in polynomial time.

Jason M. Altschuler, Enric Boix-Adserà

CoRR, 2020

The Multiplayer Colonel Blotto Game.

Enric Boix-Adserà, Benjamin L. Edelman, Siddhartha Jayanti

Proceedings of the EC '20: The 21st ACM Conference on Economics and Computation, 2020

2019

Approximating the Determinant of Well-Conditioned Matrices by Shallow Circuits.

Enric Boix-Adserà, Lior Eldar, Saeed Mehraban

CoRR, 2019

Randomized Concurrent Set Union and Generalized Wake-Up.

Siddhartha V. Jayanti, Robert E. Tarjan, Enric Boix-Adserà

Proceedings of the 2019 ACM Symposium on Principles of Distributed Computing, 2019

Sample Efficient Active Learning of Causal Trees.

Kristjan H. Greenewald, Dmitriy Katz, Karthikeyan Shanmugam, Sara Magliacane, Murat Kocaoglu, Enric Boix Adserà, Guy Bresler

Proceedings of the Advances in Neural Information Processing Systems 32: Annual Conference on Neural Information Processing Systems 2019, 2019

Subadditivity Beyond Trees and the Chi-Squared Mutual Information.

Emmanuel Abbe, Enric Boix Adserà

Proceedings of the IEEE International Symposium on Information Theory, 2019

The Average-Case Complexity of Counting Cliques in Erdős-Rényi Hypergraphs.

Enric Boix-Adserà, Matthew S. Brennan, Guy Bresler

Proceedings of the 60th IEEE Annual Symposium on Foundations of Computer Science, 2019

2018

An Information-Percolation Bound for Spin Synchronization on General Graphs.

Emmanuel Abbe, Enric Boix

CoRR, 2018