AI Paper Atlas

Research map / Architectures and efficiency / Quantization

RobuQ: Pushing DiTs to W1.58A2 via Robust Activation Quantization

Kaicheng Yang, Xun Zhang, Haotong Qin, Yucheng Lin, Kaisen Yang, Xianglong Yan, Yulun Zhang

ICML 2026 · Architectures and efficiency › Quantization · approach: Post-training quantization of ViTs · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add low-rank branch (LRB) to BitNetv2 (ablation)BitNetv2 W1.58A4 · ImageNet 256x256 W1.58A4 (cfg 1.5)FID: 41.59 -> 29.59Improved
add Hadamard transform to all linear layers on top of LRB (ablation)BitNetv2 + LRB W1.58A4 · ImageNet 256x256 W1.58A4 (cfg 1.5)FID: 29.59 -> 20.82Improved
add uniform per-token Gauss quantizer (ablation)Baseline W1.58A4 · ImageNet 256x256 W1.58A4 (cfg 1.5)FID: 20.82 -> 17.97Improved
add non-uniform per-token Gauss quantizer (ablation)Baseline W1.58A4 · ImageNet 256x256 W1.58A4 (cfg 1.5)FID: 20.82 -> 20.33No clear effect
add activation-only mixed precision (AMPN) at W1.58A3 (ablation)RobuQ w/o AMP · ImageNet 256x256 W1.58A3 (cfg 1.5)FID: 24.44 -> 21.40Improved
add activation-only mixed precision (AMPN) at W1.58A2 (ablation)RobuQ w/o AMP · ImageNet 256x256 W1.58A2 (cfg 1.5)FID: 43.31 -> 30.30Improved
raise QAT steps for metric collection from 1 to 1,000 (ablation)AMPN W1.58A2 · ImageNet 256x256 W1.58A2 (cfg 1.5)FID: 41.06 -> 30.30Improved
raise QAT steps for metric collection from 1,000 to 1,500 (ablation)AMPN W1.58A2 · ImageNet 256x256 W1.58A2 (cfg 1.5)FID: 30.30 -> 30.56No clear effect
apply RobustQuantizer W1.58A4 instead of baseline on FFHQBaseline W1.58A4 · FFHQ 256x256 (unconditional)FID: 34.32 -> 25.62Improved
apply RobuQ W1.58A2 instead of baseline on ImageNet cfg 4.0Baseline W1.58A2 · ImageNet 256x256 (cfg 4.0)FID: 75.02 -> 11.06Improved
W4A4 PTQ with RobuQ on FLUX-SchnellFLUX-Schnell · MJHQ (4 steps)FID: 18.40 -> 18.57No clear effect
W4A4 PTQ with RobuQ on alpha-Flow at NFE=1alpha-Flow-XL/2 · ImageNet-1K (NFE=1)FID: 2.58 -> 2.59No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy