Metraipolis › archive › dc_6da770ed10ff

CausalPrune: Counterfactual Causal Token Pruning for Efficient Large Vision-Language Models — arXiv:2611.00058

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.00058
License (as archived): CC0-1.0
Archived: 2026-11-03T11:40:19.262Z
SHA-256 of the archived text: 07791a5a039753a077603204298029e732ae5da7e885fd9649f1d16680fe4ba4
Stable URL of this page: https://metraipolis.com/documento/dc_6da770ed10ff

How to cite

CausalPrune: Counterfactual Causal Token Pruning for Efficient Large Vision-Language Models — arXiv:2611.00058 — arXiv (en), https://arxiv.org/abs/2611.00058 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6da770ed10ff (Metraipolis archive dc_6da770ed10ff, sha256 07791a5a039753a077603204298029e732ae5da7e885fd9649f1d16680fe4ba4)

BibTeX

@misc{dc6da770ed10ff,
  title = {CausalPrune: Counterfactual Causal Token Pruning for Efficient Large Vision-Language Models — arXiv:2611.00058},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00058},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6da770ed10ff; sha256 07791a5a039753a077603204298029e732ae5da7e885fd9649f1d16680fe4ba4},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00058. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large Vision-Language Models (LVLMs) have achieved remarkable progress in multimodal understanding, but their practical deployment is still limited by the substantial computation introduced by excessive visual tokens. Visual token pruning offers an effective solution by removing redundant visual tokens during inference. However, existing methods typically estimate token importance from correlational statistics, which cannot distinguish tokens that causally determine the model output from those spuriously associated with it. To address this issue, we propose \textbf{CausalPrune}, a counterfactual causal token pruning framework for efficient LVLMs. We formulate LVLM inference as a Structural Causal Model and introduce the Causal Token Importance Score (CTIS), which estimates the Average Treatment Effect of intervening on each visual token. To make causal estimation practical, we derive an efficient first-order approximation of CTIS and incorporate backdoor adjustment to mitigate latent cross-modal confounding. Within this causal formulation, layer-wise causal mediation decomposes token importance into direct and indirect effects through intermediate hidden states, enabling progressive pruning across transformer layers in a principled manner. Extensive experiments on various LVLM architectures, such as LLaVA-1.5 and Qwen2.5-VL, demonstrate that CausalPrune maintains comparable or even superior multimodal performance while achieving substantial reductions in FLOPs and prefill latency, consistently outperforming existing pruning methods.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2611.00058 — 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 2026-11-03T11:06:37.371Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).