Metraipolis › archive › dc_636bfb2ec909
Speed in the Blind Spot: An Interpretability Analysis of Dynamic Perception in VLMs for Autonomous Driving — arXiv:2609.37046
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.37046
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T18:40:50.478Z
- SHA-256 of the archived text
e02518d7028c3cb35496824747562f02410af68dd8ff27ba2abd7973e7825c06- Stable URL of this page
https://metraipolis.com/documento/dc_636bfb2ec909
How to cite
Speed in the Blind Spot: An Interpretability Analysis of Dynamic Perception in VLMs for Autonomous Driving — arXiv:2609.37046 — arXiv (en), https://arxiv.org/abs/2609.37046 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_636bfb2ec909 (Metraipolis archive dc_636bfb2ec909, sha256 e02518d7028c3cb35496824747562f02410af68dd8ff27ba2abd7973e7825c06)
BibTeX
@misc{dc636bfb2ec909,
title = {Speed in the Blind Spot: An Interpretability Analysis of Dynamic Perception in VLMs for Autonomous Driving — arXiv:2609.37046},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.37046},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_636bfb2ec909; sha256 e02518d7028c3cb35496824747562f02410af68dd8ff27ba2abd7973e7825c06},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.37046. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Vision-Language Models are increasingly used in autonomous-driving systems, yet their ability to recover dynamic physical state from visual input remains insufficiently characterized. We study velocity understanding as a controlled diagnostic across three tasks: surrounding-agent speed, current ego speed, and short-horizon future ego-speed proposal. On nuScenes, we evaluate open-weight general-purpose and PhysicalAI VLMs, together with the driving-oriented Alpamayo-1.5 Vision-Language-Action model, using multiple input and output formulations. We combine verbal evaluation with temporal perturbations, counterfactual ego-speed hints and linear probes of hidden representations. The tasks exhibit distinct failure modes. Surrounding-agent speed is weakly encoded in an agent-specific form, whereas current ego speed is often internally accessible but poorly verbalized: continuous probes achieve 4.7-5.8 km/h MAE compared with 10.2-16.8 km/h MAE for verbal outputs. Multiple frames provide inconsistent verbal gains to single frame inputs, and frame order is rarely exploited. Under non-optimized QLoRA, task-specific adaptation improves both task-relevant latent speed representations and verbal readout, but continuous surrounding-agent speed estimation remains weak, while most future-speed gains survive frame shuffling, indicating limited temporal grounding. Driving specialized Alpamayo-1.5 shows stronger latent representations for surrounding-agent and future ego speed, while current ego-speed decodability is comparable and substantial probe-verbal gaps remain. Thus, driving specialization can strengthen motion representations but does not guarantee stronger encoding across both scene and ego states or reliable readout. The results show that plausible planning outputs do not necessarily imply reliable recovery or temporal grounding of the underlying dynamic state. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-30 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2609.37046 — 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 2026-10-07T18:04:24.806Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).