Metraipolis › archive › dc_3d1dde668bc1

DIET: Deletion-response Expert Trimming for Video Diffusion Transformers — arXiv:2609.37829

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.37829
License (as archived)
CC0-1.0
Archived
2026-10-07T16:24:11.748Z
SHA-256 of the archived text
38c3d1d4cf26bfd0af2bc18de6c53b160afcd1e8c7f9c6e04aa7df0eb37567aa
Stable URL of this page
https://metraipolis.com/documento/dc_3d1dde668bc1

How to cite

DIET: Deletion-response Expert Trimming for Video Diffusion Transformers — arXiv:2609.37829 — arXiv (en), https://arxiv.org/abs/2609.37829 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3d1dde668bc1 (Metraipolis archive dc_3d1dde668bc1, sha256 38c3d1d4cf26bfd0af2bc18de6c53b160afcd1e8c7f9c6e04aa7df0eb37567aa)
BibTeX
@misc{dc3d1dde668bc1,
  title = {DIET: Deletion-response Expert Trimming for Video Diffusion Transformers — arXiv:2609.37829},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.37829},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3d1dde668bc1; sha256 38c3d1d4cf26bfd0af2bc18de6c53b160afcd1e8c7f9c6e04aa7df0eb37567aa},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.37829. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Video diffusion transformers (DiTs) increasingly adopt mixture-of-experts (MoE) architectures to reduce active computation, but their full expert storage remains costly. Existing one-shot pruning criteria mainly rely on static activation or routing statistics and cannot capture layer-level re-routing after expert deletion. We introduce DIET, a training-free expert pruning framework based on deletion responses. A single all-expert calibration pass records expert outputs and router states for matched conditional and unconditional tokens. Candidate deletions are then replayed from cached tensors, requiring no additional model forward passes. The resulting deletion-response signatures characterize each expert by the changes induced when it is removed. DIET selects retained experts by minimizing Overall Diversity Loss (ODL), which preserves directional coverage in signature space, and combines intra-layer local search with an inter-layer regression-guided budget search to allocate experts across layers. On LingBot-Video 30B-A3B, pruning 50% of experts (6,144 to 3,072) reduces the checkpoint from 57 GB to 30 GB and enables single-card deployment on a 48 GB GPU without fine-tuning. Under a fixed 284-case VBench protocol, the VBench Total increases from 0.7941 to 0.8115. Across tested retention budgets, DIET consistently outperforms competitive pruning baselines adapted from large language models.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.37829 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T15:54:53.660Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).