| title | Different Tasks, Different Timesteps: Temporal Task-Aware Quantization for Diffusion Language Models |
|---|---|
| authors | Ofek Avivi, Eliron Rahimi, Avi Mendelson, Chaim Baskin |
| date | 2026-10-22 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2610.31198 |
| upvotes | 19 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | Diffusion Language Models (DLMs) introduce unique quantization challenges due to their iterative denoising process. In this work, we reveal that quantization sensitivity is both taskand time-dependent, with different downstream tasks exhibiting distinct precision requirements across denoising steps. We further show that these patterns can be predicted without groundtruth labels or task-specific evaluation metrics. Building on these findings, we introduce T 2Quant, a task and time-aware mixedprecision quantization framework that uses a label-free sensitivity signal to optimally allocate activation precision across denoising steps under a fixed bit-width budget. Extensive experiments across three DLMs and six benchmarks demonstrate that T 2Quant consistently improves the average accuracy of existing quantization methods, achieving gains of up to 6.7 percentage points under the same budget. |
cite
@misc{avivi2026different,
title={Different Tasks, Different Timesteps: Temporal Task-Aware Quantization for Diffusion Language Models},
author={Ofek Avivi and Eliron Rahimi and Avi Mendelson and Chaim Baskin},
year={2026},
eprint={2610.31198},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2610.31198}
}