Breaking the Compression Barrier: Cross-Architecture Compression Boundary Learning via Reverse Regrowth
cs.LG, cs.CV
Submitted: 2026-08-17
Updated: 2026-09-15
Comments: Withdrawn by the authors following the identification of issues in the analysis and results that may materially affect the conclusions of the manuscript
Code: https://github.com/EnumaCaliber/BRIDGE
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- NIRVANA: Structured Pruning Reimagined for Large Language Model Compression
- A Comprehensive Survey on Hardware-Aware Neural Architecture Search
- Chasing Sparsity in Vision Transformers: An End-to-End Exploration
- An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Dynamic Network Surgery for Efficient DNNs
- Distilling the Knowledge in a Neural Network
- SGLP: A Similarity Guided Fast Layer Partition Pruning for Compressing Large Deep Models
- MCUNet: Tiny Deep Learning on IoT Devices
- Decoupled Weight Decay Regularization
- DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification
- MobileNetV2: Inverted Residuals and Linear Bottlenecks
- How Sparse Can We Prune A Deep Network: A Fundamental Limit Perspective
- Neural Architecture Search with Reinforcement Learning
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks