The Rescue Effect: Spatio-Semantic Early Exit Bypasses Quantization Collapse in CLIP 事件

The Rescue Effect: Spatio-Semantic Early Exit Bypasses Quantization Collapse in CLIP arXiv:2605.26415v1 Announce Type: new Abstract: Deploying Vision-Language Models on resource-constrained hardware typically requires INT8 quantization, but in joint-embedding architectures such as CLIP this introduces a failure mode distinct from quantized CNN classifiers: activation noise accumulated across transformer blocks perturbs the direction of the multimodal embedding, eroding the cosine alignment on w