Advait Gupta, Sachin Konan, Tom Yin, Gabriel Sarch, Zhuang Liu · 2026 · arXiv
DOI: 10.48550/arXiv.2610.31494 · open-access PDF
Los modelos de lenguaje de visión (VLM) pueden realizar una amplia gama de tareas en diferentes configuraciones, pero su rendimiento en contextos de agencia sigue siendo mal comprendido. Aunque las evaluaciones en línea en simuladores y en entornos del mundo real son esenciales para medir el rendimiento del agente interactivo, son costosos y difíciles de escalar. Los benchmarks de VQA en línea complementarios pueden permitir comparaciones más rápidas y fiables entre dominios de agencia, pero los benchmarks existentes siguen siendo limitados y fragmentados.
Machine translation (Meta M2M100); the English original follows.
Vision-language models (VLMs) can perform a broad range of tasks across diverse settings, yet their performance in agentic contexts remains poorly understood. Although online evaluations in simulators and real-world environments are essential for measuring interactive agent performance, they are costly and challenging to scale. Complementary offline VQA benchmarks can enable faster and more reliable comparisons across agentic domains, yet existing benchmarks remain limited and fragmented. To address this gap, we introduce AgentVQA, an offline benchmark for evaluating VLMs on realistic agentic
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.31494, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange