CALR: Corrective Adaptive Low-Rank Decomposition for Efficient Large Language Model Layer Compression

TL;DR

CALR introduces a novel low-rank decomposition method with a learnable correction module to effectively compress large language models while preserving their performance, enabling more practical deployment in resource-limited settings.

Contribution

It proposes CALR, a two-part compression technique combining SVD with a learnable correction module to better retain model functionality after compression.

Findings

Reduces model size by up to 51.77%.

Retains up to 90.42% of original performance.

Outperforms existing compression methods.

Abstract

Large Language Models (LLMs) present significant deployment challenges due to their immense size and computational requirements. Model compression techniques are essential for making these models practical for resource-constrained environments. A prominent compression strategy is low-rank factorization via Singular Value Decomposition (SVD) to reduce model parameters by approximating weight matrices. However, standard SVD focuses on minimizing matrix reconstruction error, often leading to a substantial loss of the model's functional performance. This performance degradation occurs because existing methods do not adequately correct for the functional information lost during compression. To address this gap, we introduce Corrective Adaptive Low-Rank Decomposition (CALR), a two-component compression approach. CALR combines a primary path of SVD-compressed layers with a parallel, learnable, low-rank corrective module that is explicitly trained to recover the functional residual error. Our experimental evaluation on SmolLM2-135M, Qwen3-0.6B, and Llama-3.2-1B, demonstrates that CALR can reduce parameter counts by 26.93% to 51.77% while retaining 59.45% to 90.42% of the original model's performance, consistently outperforming LaCo, ShortGPT, and LoSparse. CALR's success shows that treating functional information loss as a learnable signal is a highly effective compression paradigm. This approach enables the creation of significantly smaller, more efficient LLMs, advancing their accessibility and practical deployment in real-world applications.