Can Less Precise Be More Reliable? A Systematic Evaluation of Quantization's Impact on CLIP Beyond Accuracy
This work addresses the problem of deploying efficient and reliable vision-language models for safety-related tasks, offering insights that challenge the conventional efficiency-performance trade-off, though it is incremental in exploring quantization effects.
The study systematically evaluated the impact of quantization on CLIP models, finding that quantization can improve calibration for underconfident models and enhance out-of-distribution detection even when calibration degrades, with specific methods achieving simultaneous gains in accuracy, calibration, and robustness.
The powerful zero-shot generalization capabilities of vision-language models (VLMs) like CLIP have enabled new paradigms for safety-related tasks such as out-of-distribution (OOD) detection. However, additional aspects crucial for the computationally efficient and reliable deployment of CLIP are still overlooked. In particular, the impact of quantization on CLIP's performance beyond accuracy remains underexplored. This work presents a large-scale evaluation of quantization on CLIP models, assessing not only in-distribution accuracy but a comprehensive suite of reliability metrics and revealing counterintuitive results driven by pre-training source. We demonstrate that quantization consistently improves calibration for typically underconfident pre-trained models, while often degrading it for overconfident variants. Intriguingly, this degradation in calibration does not preclude gains in other reliability metrics; we find that OOD detection can still improve for these same poorly calibrated models. Furthermore, we identify specific quantization-aware training (QAT) methods that yield simultaneous gains in zero-shot accuracy, calibration, and OOD robustness, challenging the view of a strict efficiency-performance trade-off. These findings offer critical insights for navigating the multi-objective problem of deploying efficient, reliable, and robust VLMs by utilizing quantization beyond its conventional role.