Search papers, labs, and topics across Lattice.
This paper introduces QuantWAMs, a post-training quantization framework specifically designed for World Action Models (WAMs) that addresses inefficiencies in existing methods by aligning quantization with the model's calibration context. The framework employs three innovative strategies, including shared-basis outlier calibration and co-training-objective saliency, to enhance the precision and efficiency of WAMs during deployment. Evaluation results demonstrate that QuantWAMs achieves significant memory reduction and speed improvements while maintaining high accuracy in both simulation and real-robot manipulation tasks.
QuantWAMs slashes memory usage to 29% of FP16 while delivering up to 1.6x speedups in real-robot tasks without sacrificing performance.
World Action Models (WAMs) jointly predict future observations and actions, but their iterative denoising and closed-loop execution make efficient deployment costly. Existing post-training quantization (PTQ) methods are poorly suited to WAMs because they rely on open-loop objectives, homogeneous model assumptions, and calibration distributions that do not reflect deployment. We present QuantWAMs, a PTQ framework that aligns quantization decisions with the calibration context defined by model structure, rollout distribution, and task objective. QuantWAMs introduces three strategies: shared-basis outlier calibration, which pools activation evidence only across coordinate-compatible modules; co-training-objective saliency, which computes empirical-Fisher scores from the joint video--action gradient and assigns weight precision at a calibration-stable layer granularity; and fixed-intervention rollout auditing, which revises denoising-step protection schedules using reachable closed-loop states without changing the precision budget. We evaluate QuantWAMs on Fast-WAM and LingBot-VA across RoboTwin 2.0, LIBERO, and real-robot manipulation with an AgiBot G2. Under a W4A4-dominant setting, the reported simulation means differ from FP16 by 0.2--0.7 percentage points. Real-robot trials further establish deployment feasibility on three manipulation tasks. For the targeted video and action blocks, QuantWAMs reduces peak weight-and-activation memory to about 29\% of FP16 and provides 1.4--1.6$\times$ block-level speedups.
