Cover Image

Compress Then Adapt? No, Do It Together via Task-aware Union of Subspaces

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Ge, Jingze, Liu, Yun, Geng, Xue, Dong, Wanqi, Mark, Wang Zhe, Wu, Min, Yang, Xulei
Format:	Preprint
Published:	2026
Subjects:	Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

CERSA: Cumulative Energy-Retaining Subspace Adaptation for Memory-Efficient Fine-Tuning
by: Ge, Jingze, et al.
Published: (2026)

QuadBEV: An Efficient Quadruple-Task Perception Framework via Bird's-Eye-View Representation
by: Li, Yuxin, et al.
Published: (2024)

Wonderful Matrices: More Efficient and Effective Architecture for Language Modeling Tasks
by: Shi, Jingze, et al.
Published: (2024)

Information-Theoretic Optimization for Task-Adapted Compressed Sensing Magnetic Resonance Imaging
by: Peng, Xinyu, et al.
Published: (2026)

Mixture-of-Subspaces in Low-Rank Adaptation
by: Wu, Taiqiang, et al.
Published: (2024)

Purifying Task Vectors in Knowledge-Aware Subspace for Model Merging
by: An, Bang, et al.
Published: (2025)

Acquiring and Adapting Priors for Novel Tasks via Neural Meta-Architectures
by: Babu, Sudarshan
Published: (2025)

Model Merging in the Essential Subspace
by: Li, Longhua, et al.
Published: (2026)

From Algorithm to Hardware: A Survey on Efficient and Safe Deployment of Deep Neural Networks
by: Geng, Xue, et al.
Published: (2024)

Wonderful Matrices: Combining for a More Efficient and Effective Foundation Model Architecture
by: Shi, Jingze, et al.
Published: (2024)

AAT: Adapting Audio Transformer for Various Acoustics Recognition Tasks
by: Liang, Yun, et al.
Published: (2024)

Low-Rank Compression of Pretrained Models via Randomized Subspace Iteration
by: Pourkamali-Anaraki, Farhad
Published: (2026)

Beyond Variance: Knowledge-Aware LLM Compression via Fisher-Aligned Subspace Diagnostics
by: Shihab, Ibne Farabi, et al.
Published: (2026)

Accelerating Data Generation for Neural Operators via Krylov Subspace Recycling
by: Wang, Hong, et al.
Published: (2024)

TreeSynth: Synthesizing Diverse Data from Scratch via Tree-Guided Subspace Partitioning
by: Wang, Sheng, et al.
Published: (2025)

Domain Generalizable Knowledge Tracing via Concept Aggregation and Relation-Based Attention
by: Xie, Yuquan, et al.
Published: (2024)

Hyper-Compression: Model Compression via Hyperfunction
by: Fan, Fenglei, et al.
Published: (2024)

MedUnifier: Unifying Vision-and-Language Pre-training on Medical Data with Vision Generation Task using Discrete Visual Representations
by: Zhang, Ziyang, et al.
Published: (2025)

From Long to Lean: Performance-aware and Adaptive Chain-of-Thought Compression via Multi-round Refinement
by: Yan, Jianzhi, et al.
Published: (2025)

How Do Images Align and Complement LiDAR? Towards a Harmonized Multi-modal 3D Panoptic Segmentation
by: Pan, Yining, et al.
Published: (2025)

Task-Aware Encoder Control for Deep Video Compression
by: Ge, Xingtong, et al.
Published: (2024)

Do Retrieval-Augmented Language Models Adapt to Varying User Needs?
by: Wu, Peilin, et al.
Published: (2025)

Birds of a Feather Flock Together: Background-Invariant Representations via Linear Structure in VLMs
by: Zaazou, Youssef, et al.
Published: (2026)

SceneAdapt: Scene-aware Adaptation of Human Motion Diffusion
by: Cho, Jungbin, et al.
Published: (2025)

Enhancing Delta Compression in LLMs via SVD-based Quantization Error Minimization
by: Xiong, Boya, et al.
Published: (2025)

ADAPT: Actively Discovering and Adapting to Preferences for any Task
by: Patel, Maithili, et al.
Published: (2025)

Do not Abstain! Identify and Solve the Uncertainty
by: Liu, Jingyu, et al.
Published: (2025)

Learning Multi-Indicator Weights for Data Selection: A Joint Task-Model Adaptation Framework with Efficient Proxies
by: Song, Jingze, et al.
Published: (2026)

OpenRFT: Adapting Reasoning Foundation Model for Domain-specific Tasks with Reinforcement Fine-Tuning
by: Zhang, Yuxiang, et al.
Published: (2024)

A Guardrail for Safety Preservation: When Safety-Sensitive Subspace Meets Harmful-Resistant Null-Space
by: Zhang, Bingjie, et al.
Published: (2025)

MALLOC: Benchmarking the Memory-aware Long Sequence Compression for Large Sequential Recommendation
by: Yu, Qihang, et al.
Published: (2026)

Contribution-aware Token Compression for Efficient Video Understanding via Reinforcement Learning
by: Ma, Yinchao, et al.
Published: (2026)

Scaling Synthetic Task Generation for Agents via Exploration
by: Ramrakhya, Ram, et al.
Published: (2025)

OpenDeception: Learning Deception and Trust in Human-AI Interaction via Multi-Agent Simulation
by: Wu, Yichen, et al.
Published: (2025)

Conjunction Subspaces Test for Conformal and Selective Classification
by: He, Zengyou, et al.
Published: (2024)

Shared Parameter Subspaces and Cross-Task Linearity in Emergently Misaligned Behavior
by: Arturi, Daniel Aarao Reis, et al.
Published: (2025)

Toward Scalable Terminal Task Synthesis via Skill Graphs
by: Fan, Zhiyuan, et al.
Published: (2026)

Effective and Efficient One-pass Compression of Speech Foundation Models Using Sparsity-aware Self-pinching Gates
by: Xu, Haoning, et al.
Published: (2025)

C$^3$TG: Conflict-aware, Composite, and Collaborative Controlled Text Generation
by: Li, Yu, et al.
Published: (2025)

AdaptFuse: Training-Free Sequential Preference Learning via Externalized Bayesian Inference
by: Lin, Fangzhou, et al.
Published: (2026)