paper-with-me

홈 › Papers

Universal Hypernetworks for Arbitrary Models

2026-04-02 · Xuanfeng Zhou arxiv

Conventional hypernetworks are typically engineered around a specific base-model parameterization, so changing the target architecture often entails redesigning the hypernetwork and retraining it from scratch. We introduce the \emph{Universal Hypernetwork} (UHN), a fixed-architecture generator that predicts weights from deterministic parameter, architecture, and task descriptors. This descriptor-based formulation decouples the generator architecture from target-network parameterization, so one generator can instantiate heterogeneous models across the tested architecture and task families. Our empirical claims are threefold: (1) one fixed UHN remains competitive with direct training across vision, graph, text, and formula-regression benchmarks; (2) the same UHN supports both multi-model generalization within a family and multi-task learning across heterogeneous models; and (3) UHN enables stable recursive generation with up to three intermediate generated UHNs before the final base model. Our code is available at https://github.com/Xuanfeng-Zhou/UHN.

📄 PDF Abstract BibTeX arXiv:2604.02215

Code (0)

등록된 구현이 없습니다.

Tasks

Multi-Task Learning

Similar Papers 제목 키워드 기반

Learning Maximal Safe Sets Using Hypernetworks for MPC-based Local Trajectory Planning in Unknown Environments

2024-10-26 · Bojan Derajić, Mohamed-Khalil Bouzidi, Sebastian Bernhard, Wolfgang Hönig

This paper presents a novel learning-based approach for online estimation of maximal safe sets for local trajectory planning in unknown static environments. The neural representation of a set is used as the terminal set …

Model Predictive ControlMotion PlanningTrajectory Planning

HyperNetworks

2016-09-27 · David Ha, Andrew Dai, Quoc V. Le

This work explores hypernetworks: an approach of using a one network, also known as a hypernetwork, to generate the weights for another network. Hypernetworks provide an abstraction that is similar to what is found in na…

Handwriting generationLanguage ModellingMachine TranslationTranslation

Distilling Morphology-Conditioned Hypernetworks for Efficient Universal Morphology Control

2024-02-09 · Zheng Xiong, Risto Vuorio, Jacob Beck, Matthieu Zimmer 외

Learning a universal policy across different robot morphologies can significantly improve learning efficiency and enable zero-shot generalization to unseen morphologies. However, learning a highly performant universal po…

Zero-shot Generalization

Effect of Optimizer, Initializer, and Architecture of Hypernetworks on Continual Learning from Demonstration

2023-12-31 · Sayantan Auddy, Sebastian Bergner, Justus Piater

In continual learning from demonstration (CLfD), a robot learns a sequence of real-world motion skills continually from human demonstrations. Recently, hypernetworks have been successful in solving this problem. In this …

Continual Learning

Hypernetwork approach to Bayesian MAML

2022-10-06 · Piotr Borycki, Piotr Kubacki, Marcin Przewięźlikowski, Tomasz Kuśmierczyk 외

The main goal of Few-Shot learning algorithms is to enable learning from small amounts of data. One of the most popular and elegant Few-Shot learning approaches is Model-Agnostic Meta-Learning (MAML). The main idea behin…

Few-Shot LearningMeta-Learning