Imagen de Portada

When Models Don't Collapse: On the Consistency of Iterative MLE

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Barzilai, Daniel, Shamir, Ohad
Formato:	Preprint
Publicado:	2025
Materias:	Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Limitations of SGD for Multi-Index Models Beyond Statistical Queries
por: Barzilai, Daniel, et al.
Publicado: (2026)

Generalization in Kernel Regression Under Realistic Assumptions
por: Barzilai, Daniel, et al.
Publicado: (2023)

Simple Relative Deviation Bounds for Covariance and Gram Matrices
por: Barzilai, Daniel, et al.
Publicado: (2024)

Beyond Benign Overfitting in Nadaraya-Watson Interpolators
por: Barzilai, Daniel, et al.
Publicado: (2025)

Are Convex Optimization Curves Convex?
por: Barzilai, Guy, et al.
Publicado: (2025)

Gradient Descent's Last Iterate is Often (slightly) Suboptimal
por: Kornowski, Guy, et al.
Publicado: (2026)

REED-VAE: RE-Encode Decode Training for Iterative Image Editing with Diffusion Models
por: Almog, Gal, et al.
Publicado: (2025)

Hardness of Learning Fixed Parities with Neural Networks
por: Shoshani, Itamar, et al.
Publicado: (2025)

An Algorithm with Optimal Dimension-Dependence for Zero-Order Nonsmooth Nonconvex Stochastic Optimization
por: Kornowski, Guy, et al.
Publicado: (2023)

On the Complexity of Finding Small Subgradients in Nonsmooth Optimization
por: Kornowski, Guy, et al.
Publicado: (2022)

Open Problem: Anytime Convergence Rate of Gradient Descent
por: Kornowski, Guy, et al.
Publicado: (2024)

Experts Don't Cheat: Learning What You Don't Know By Predicting Pairs
por: Johnson, Daniel D., et al.
Publicado: (2024)

LoRA and Privacy: When Random Projections Help (and When They Don't)
por: Hu, Yaxi, et al.
Publicado: (2026)

When Is Compositional Reasoning Learnable from Verifiable Rewards?
por: Barzilai, Daniel, et al.
Publicado: (2026)

Logarithmic Width Suffices for Robust Memorization
por: Egosi, Amitsour, et al.
Publicado: (2025)

Implicit Regularization Towards Rank Minimization in ReLU Networks
por: Timor, Nadav, et al.
Publicado: (2022)

The Oracle Complexity of Simplex-based Matrix Games
por: Kornowski, Guy, et al.
Publicado: (2024)

On the Hardness of Meaningful Local Guarantees in Nonsmooth Nonconvex Optimization
por: Kornowski, Guy, et al.
Publicado: (2024)

From Tempered to Benign Overfitting in ReLU Neural Networks
por: Kornowski, Guy, et al.
Publicado: (2023)

mHC-lite: You Don't Need 20 Sinkhorn-Knopp Iterations
por: Yang, Yongyi, et al.
Publicado: (2026)

When Respondents Don't Care Anymore: Identifying the Onset of Careless Responding
por: Welz, Max, et al.
Publicado: (2023)

What LLMs Think When You Don't Tell Them What to Think About?
por: Kwon, Yongchan, et al.
Publicado: (2026)

If You Don't Understand It, Don't Use It: Eliminating Trojans with Filters Between Layers
por: Hernandez, Adriano
Publicado: (2024)

MLE-Smith: Scaling MLE Tasks with Automated Multi-Agent Pipeline
por: Qiang, Rushi, et al.
Publicado: (2025)

Depth Separation in Norm-Bounded Infinite-Width Neural Networks
por: Parkinson, Suzanna, et al.
Publicado: (2024)

Don't Freeze, Don't Crash: Extending the Safe Operating Range of Neural Navigation in Dense Crowds
por: Zhang, Jiefu, et al.
Publicado: (2026)

Don't Throw Away Your Beams: Improving Consistency-based Uncertainties in LLMs via Beam Search
por: Fadeeva, Ekaterina, et al.
Publicado: (2025)

Position: Don't be Afraid of Over-Smoothing And Over-Squashing
por: Kormann, Niklas, et al.
Publicado: (2026)

Grow, Don't Overwrite: Fine-tuning Without Forgetting
por: Adila, Dyah, et al.
Publicado: (2026)

Reasoning Models Don't Just Think Longer, They Move Differently
por: Gjølbye, Anders, et al.
Publicado: (2026)

I Don't Know: Explicit Modeling of Uncertainty with an [IDK] Token
por: Cohen, Roi, et al.
Publicado: (2024)

Don't Run with Scissors: Pruning Breaks VLA Models but They Can Be Recovered
por: Jabbour, Jason, et al.
Publicado: (2025)

Why Do Some Language Models Fake Alignment While Others Don't?
por: Sheshadri, Abhay, et al.
Publicado: (2025)

Don't be so negative! Score-based Generative Modeling with Oracle-assisted Guidance
por: Naderiparizi, Saeid, et al.
Publicado: (2023)

Reasoning Models Don't Always Say What They Think
por: Chen, Yanda, et al.
Publicado: (2025)

You Don't Need Domain-Specific Data Augmentations When Scaling Self-Supervised Learning
por: Moutakanni, Théo, et al.
Publicado: (2024)

We Still Don't Understand High-Dimensional Bayesian Optimization
por: Doumont, Colin, et al.
Publicado: (2025)

Don't Stop Me Now: Embedding Based Scheduling for LLMs
por: Shahout, Rana, et al.
Publicado: (2024)

Sample, Don't Search: Rethinking Test-Time Alignment for Language Models
por: Faria, Gonçalo, et al.
Publicado: (2025)

Don't Forget Imagination!
por: Vityaev, Evgenii E., et al.
Publicado: (2025)