Imagen de Portada

Emergence of Frontier Superposition: Möbius attractor and Cascade Supervision

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Gu, Hongyu, Fu, Jingwen
Formato:	Preprint
Publicado:	2026
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

Superposition in Graph Neural Networks
por: Pertl, Lukas, et al.
Publicado: (2025)

Sparsity and Superposition in Mixture of Experts
por: Chaudhari, Marmik, et al.
Publicado: (2025)

Mathematical Models of Computation in Superposition
por: Hänni, Kaarel, et al.
Publicado: (2024)

On Implications of Scaling Laws on Feature Superposition
por: Katta, Pavan
Publicado: (2024)

Superposition Prompting: Improving and Accelerating Retrieval-Augmented Generation
por: Merth, Thomas, et al.
Publicado: (2024)

Expanding Expressivity in Transformer Models with MöbiusAttention
por: Halacheva, Anna-Maria, et al.
Publicado: (2024)

Understanding Emergent Misalignment via Feature Superposition Geometry
por: Minegishi, Gouki, et al.
Publicado: (2026)

RanDeS: Randomized Delta Superposition for Multi-Model Compression
por: Zhou, Hangyu, et al.
Publicado: (2025)

Discrete, compositional, and symbolic representations through attractor dynamics
por: Nam, Andrew, et al.
Publicado: (2023)

Superposition as Lossy Compression: Measure with Sparse Autoencoders and Connect to Adversarial Vulnerability
por: Bereska, Leonard, et al.
Publicado: (2025)

Policy Regularized Distributionally Robust Markov Decision Processes with Linear Function Approximation
por: Gu, Jingwen, et al.
Publicado: (2025)

Superposition Yields Robust Neural Scaling
por: Liu, Yizhou, et al.
Publicado: (2025)

Spectral Superposition: A Theory of Feature Geometry
por: Ivanov, Georgi, et al.
Publicado: (2026)

The Geometry of Multi-Task Grokking: Transverse Instability, Superposition, and Weight Decay Phase Structure
por: Xu, Yongzhong
Publicado: (2026)

Geometric Limits of Knowledge Distillation: A Minimum-Width Theorem via Superposition Theory
por: Sarkar, Nilesh, et al.
Publicado: (2026)

On the dimension of pullback attractors in recurrent neural networks
por: Fadera, Muhammed
Publicado: (2025)

LacMaterial: Large Language Models as Analogical Chemists for Materials Discovery
por: Guo, Hongyu
Publicado: (2025)

Foundations and Frontiers of Graph Learning Theory
por: Huang, Yu, et al.
Publicado: (2024)

Frontier Models are Capable of In-context Scheming
por: Meinke, Alexander, et al.
Publicado: (2024)

Provably Extracting the Features from a General Superposition
por: Liu, Allen
Publicado: (2025)

Probing for Representation Manifolds in Superposition
por: Modell, Alexander
Publicado: (2026)

Structured Progressive Knowledge Activation for LLM-Driven Neural Architecture Search
por: Liu, Zhen, et al.
Publicado: (2026)

A Unified Knowledge-Distillation and Semi-Supervised Learning Framework to Improve Industrial Ads Delivery Systems
por: Eghbalzadeh, Hamid, et al.
Publicado: (2025)

Mixture of Experts based Multi-task Supervise Learning from Crowds
por: Han, Tao, et al.
Publicado: (2024)

CHGNN: A Semi-Supervised Contrastive Hypergraph Learning Network
por: Song, Yumeng, et al.
Publicado: (2023)

Test Time Training for Supervised Causal Learning
por: Deng, Zizhen, et al.
Publicado: (2026)

Superposition unifies power-law training dynamics
por: Chen, Zixin Jessie, et al.
Publicado: (2026)

Supervised Contrastive Learning based Dual-Mixer Model for Remaining Useful Life Prediction
por: Fu, En, et al.
Publicado: (2024)

Confabulation dynamics in a reservoir computer: Filling in the gaps with untrained attractors
por: O'Hagan, Jack, et al.
Publicado: (2025)

Compute Requirements for Algorithmic Innovation in Frontier AI Models
por: Barnett, Peter
Publicado: (2025)

Combinatorial Creativity: A New Frontier in Generalization Abilities
por: Schapiro, Samuel, et al.
Publicado: (2025)

Self-Supervised Learning of Disentangled Representations for Multivariate Time-Series
por: Chang, Ching, et al.
Publicado: (2024)

Everything Everywhere All at Once: LLMs can In-Context Learn Multiple Tasks in Superposition
por: Xiong, Zheyang, et al.
Publicado: (2024)

Adversarial Attacks Leverage Interference Between Features in Superposition
por: Stevinson, Edward, et al.
Publicado: (2025)

Beyond the Frontier: Stochastic Backtracking for Efficient Test-Time Scaling
por: Tran, Dao, et al.
Publicado: (2026)

PACED: Distillation and On-Policy Self-Distillation at the Frontier of Student Competence
por: Xu, Yuanda, et al.
Publicado: (2026)

Expert Upcycling: Shifting the Compute-Efficient Frontier of Mixture-of-Experts
por: Dwivedi, Chaitanya, et al.
Publicado: (2026)

Frontier Large Language Models Rival State-of-the-Art Planners
por: Corrêa, Augusto B., et al.
Publicado: (2025)

Exploring the Frontiers of LLMs in Psychological Applications: A Comprehensive Review
por: Ke, Luoma, et al.
Publicado: (2024)

Estimating Worst-Case Frontier Risks of Open-Weight LLMs
por: Wallace, Eric, et al.
Publicado: (2025)