E-PMQ: Expert-Guided Post-Merge Quantization with Merged-Weight Anchoring
Quick Answer
E-PMQ enhances post-merge quantization, improving 4-bit GPTQ performance from 65.0% to 73.6% on CLIP-ViT-B/32 and from 34.8% to 76.7% on 20-task CLIP-ViT-L/14.
Quick Take
This expert-guided framework stabilizes calibration and integrates multiple models efficiently for low-resource deployment.
Key Points
- E-PMQ uses expert weights for layer-wise calibration in post-merge quantization.
- Merged-weight anchoring stabilizes calibration and preserves merged model behavior.
- Significant performance improvements observed across multiple task settings.
- Demonstrates effective low-bit deployment for neural networks.
- Addresses quantization and merging deviations in model performance.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Low-resource deployment constraints have made model quantization essential for deploying neural networks while preserving performance. Meanwhile, model merging has become an increasingly practical low-resource strategy for integrating multiple task- or domain-specialized experts into a single model without joint training or multi-model serving. Together, quantization and model merging enable an efficient low-resource deployment pipeline by integrating multiple experts into one low-bit model. We formulate this setting as Post-Merge Quantization (PMQ). We show that directly applying post-training quantization (PTQ) to a merged model is unreliable because two distinct deviations are coupled: the quantization deviation introduced by low-bit reconstruction and the expert-relative merging deviation inherited from model merging. To mitigate these deviations, we propose E-PMQ, an expert-guided PMQ framework that uses source expert weights to provide expert- guided output targets during layer-wise calibration, together with merged-weight anchoring to stabilize the calibration and preserve the integrated behavior of the merged model. On CLIP-ViT-B/32 eight-task merging, E-PMQ improves 4-bit GPTQ from 65.0% to 73.6% under Task Arithmetic and from 69.1% to 74.8% under TIES-Merging. On harder settings, E-PMQ improves GPTQ from 34.8% to 76.7% on 20-task CLIP-ViT-L/14 and from 78.26% to 83.34% on FLAN-T5- base GLUE. These results demonstrate that E-PMQ enables effective post-merge quantization and low-bit deployment.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2605.16882 [cs.CL] |
| (or arXiv:2605.16882v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2605.16882 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Wenjun Wang [view email]
[v1]
Sat, 16 May 2026 08:44:36 UTC (206 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →TriAgent: Divergence-Aware Committees for Cost-Efficient Financial Sentiment Analysis
TriAgent introduces a cost-efficient multi-agent system for financial sentiment analysis, combining VADER, FinBERT, and Qwen2.5. It achieves an F1 score of ~0.87 with significant savings of $9.3M/year at a 10M-user scale compared to GPT-4o-mini, while also detecting hallucinations with an AUC of 0.90.