Metraipolis › archive › dc_7b30309ea092

EviRover: Reinforcing Agentic Perception Beyond a Glance — arXiv:2609.40230

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.40230
License (as archived): CC0-1.0
Archived: 2026-10-07T12:46:16.763Z
SHA-256 of the archived text: 24a32c0931bc4bcc8939c3d2f37af27074e80535359ce2f0b4ca9c708a748983
Stable URL of this page: https://metraipolis.com/documento/dc_7b30309ea092

How to cite

EviRover: Reinforcing Agentic Perception Beyond a Glance — arXiv:2609.40230 — arXiv (en), https://arxiv.org/abs/2609.40230 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7b30309ea092 (Metraipolis archive dc_7b30309ea092, sha256 24a32c0931bc4bcc8939c3d2f37af27074e80535359ce2f0b4ca9c708a748983)

BibTeX

@misc{dc7b30309ea092,
  title = {EviRover: Reinforcing Agentic Perception Beyond a Glance — arXiv:2609.40230},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.40230},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7b30309ea092; sha256 24a32c0931bc4bcc8939c3d2f37af27074e80535359ce2f0b4ca9c708a748983},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.40230. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Visual perception is conventionally formulated as a one-shot prediction from a single glance at the image, under the assumption that the image content and the model's parametric knowledge suffice to resolve the query. This assumption often fails in real-world scenarios that hinge on fine-grained visual details or require knowledge-intensive and up-to-date information. We term such cases \textit{perception under insufficient evidence} and formulate perception as an agentic process that can obtain information beyond a single glance. To address the absence of data for this setting, we design two dedicated data generation pipelines, yielding EviRover-SFT-5K and EviRover-RL-12K for training. We further construct EviLens, a human-verified benchmark comprising 688 instances across five perception categories. Building on these data, we present EviRover, to our knowledge the first perception agent explicitly trained to resolve perceptual queries through interaction, using supervised fine-tuning followed by agentic reinforcement learning. Experiments show that the 4B EviRover outperforms its backbone by 30 points on average on EviLens, reaching performance comparable to advanced proprietary models. The gains transfer beyond EviLens to WebEyes, conventional perception benchmarks, and general multimodal benchmarks, including a 15-point improvement on BrowseComp-VL. All code, models, and data are released.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.40230 — 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 2026-10-07T12:14:14.237Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).