Imagen de Portada

On the Mathematical Impossibility of Safe Universal Approximators

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autor principal:	Yao, Jasper
Formato:	Preprint
Publicado:	2025
Materias:	Machine Learning Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

AlgoSelect: Universal Algorithm Selection via the Comb Operator
por: Yao, Jasper
Publicado: (2025)

Universal Approximation with Softmax Attention
por: Hu, Jerry Yao-Chieh, et al.
Publicado: (2025)

Impossibility Theorems for Feature Attribution
por: Bilodeau, Blair, et al.
Publicado: (2022)

Attention Mechanism, Max-Affine Partition, and Universal Approximation
por: Liu, Hude, et al.
Publicado: (2025)

The Impossibility of Inverse Permutation Learning in Transformer Models
por: Alur, Rohan, et al.
Publicado: (2025)

Mathematics and Coding are Universal AI Benchmarks
por: Chojecki, Przemyslaw
Publicado: (2025)

Mission: Impossible Language Models
por: Kallini, Julie, et al.
Publicado: (2024)

The Alignment Trap: Complexity Barriers
por: Yao, Jasper
Publicado: (2025)

The Impossibility Triangle of Long-Context Modeling
por: Zhou, Yan
Publicado: (2026)

Approximate Model-Based Shielding for Safe Reinforcement Learning
por: Goodall, Alexander W., et al.
Publicado: (2023)

SafeSeek: Universal Attribution of Safety Circuits in Language Models
por: Yu, Miao, et al.
Publicado: (2026)

Universal Approximation Theorem of Deep Q-Networks
por: Qi, Qian
Publicado: (2025)

Universal Approximation Theorem for a Single-Layer Transformer
por: Gumaan, Esmail
Publicado: (2025)

Mission Impossible: A Statistical Perspective on Jailbreaking LLMs
por: Su, Jingtong, et al.
Publicado: (2024)

Safe RLHF Beyond Expectation: Stochastic Dominance for Universal Spectral Risk Control
por: Chittepu, Yaswanth, et al.
Publicado: (2026)

A General Method for Proving Networks Universal Approximation Property
por: Wang, Wei
Publicado: (2025)

Dynamic Universal Approximation Theory: Foundations for Parallelism in Neural Networks
por: Wang, Wei, et al.
Publicado: (2024)

Stacked Universal Successor Feature Approximators for Safety in Reinforcement Learning
por: Cannon, Ian, et al.
Publicado: (2024)

Safe Deployment of Offline Reinforcement Learning via Input Convex Action Correction
por: Durkin, Alex, et al.
Publicado: (2025)

From Universal Approximation Theorem to Tropical Geometry of Multi-Layer Perceptrons
por: Chu, Yi-Shan, et al.
Publicado: (2025)

HardNet: Hard-Constrained Neural Networks with Universal Approximation Guarantees
por: Min, Youngjae, et al.
Publicado: (2024)

Transformer Approximations from ReLUs
por: Hu, Jerry Yao-Chieh, et al.
Publicado: (2026)

Constraint-Conditioned Policy Optimization for Versatile Safe Reinforcement Learning
por: Yao, Yihang, et al.
Publicado: (2023)

VAR-MATH: Probing True Mathematical Reasoning in LLMS via Symbolic Multi-Instance Benchmarks
por: Yao, Jian, et al.
Publicado: (2025)

SafeAR: Safe Algorithmic Recourse by Risk-Aware Policies
por: Wu, Haochen, et al.
Publicado: (2023)

Safe: Enhancing Mathematical Reasoning in Large Language Models via Retrospective Step-aware Formal Verification
por: Liu, Chengwu, et al.
Publicado: (2025)

SafeAdapt: Provably Safe Policy Updates in Deep Reinforcement Learning
por: Anisimov, Maksim, et al.
Publicado: (2026)

UniCoMTE: A Universal Counterfactual Framework for Explaining Time-Series Classifiers on ECG Data
por: Li, Justin, et al.
Publicado: (2025)

Advancing Constrained Monotonic Neural Networks: Achieving Universal Approximation Beyond Bounded Activations
por: Sartor, Davide, et al.
Publicado: (2025)

OptEx: Expediting First-Order Optimization with Approximately Parallelized Iterations
por: Shu, Yao, et al.
Publicado: (2024)

Safe RLHF-V: Safe Reinforcement Learning from Multi-modal Human Feedback
por: Ji, Jiaming, et al.
Publicado: (2025)

SafeMIL: Learning Offline Safe Imitation Policy from Non-Preferred Trajectories
por: Burnwal, Returaj, et al.
Publicado: (2025)

Revisiting Safe Exploration in Safe Reinforcement learning
por: Eckel, David, et al.
Publicado: (2024)

Projection Methods for Operator Learning and Universal Approximation
por: Zappala, Emanuele
Publicado: (2024)

Safe Flow Q-Learning: Offline Safe Reinforcement Learning with Reachability-Based Flow Policies
por: Tayal, Mumuksh, et al.
Publicado: (2026)

SafeAuto: Knowledge-Enhanced Safe Autonomous Driving with Multimodal Foundation Models
por: Zhang, Jiawei, et al.
Publicado: (2025)

Impossible Distillation: from Low-Quality Model to High-Quality Dataset & Model for Summarization and Paraphrasing
por: Jung, Jaehun, et al.
Publicado: (2023)

Universal In-Context Approximation By Prompting Fully Recurrent Models
por: Petrov, Aleksandar, et al.
Publicado: (2024)

Implicit Safe Set Algorithm for Provably Safe Reinforcement Learning
por: Zhao, Weiye, et al.
Publicado: (2024)

Probabilistic Shielding for Safe Reinforcement Learning
por: Court, Edwin Hamel-De le, et al.
Publicado: (2025)