LATIDIA · Robótica

Pensemos en dos pasos: mitigar el sesgo de concordancia en los MLLM mediante la verificación autofundamentada

arXiv:2507.11662v4 Tipo de anuncio: sustitución cruzada Resumen: Los verificadores (funciones que asignan recompensas al comportamiento de los agentes) han sido clave para el progreso de la IA en matemáticas, programación y juegos. Sin embargo, extender los beneficios a dominios sin

La noticia

arXiv:2507.11662v4 Announce Type: replace-cross Abstract: Verifiers--functions assigning rewards to agent behavior--have been key to AI progress in math, code, and games. However, extending gains to domains without clear-cut success criteria (e.g., computer use) remains a challenge: while humans can recognize desired outcomes, translating this intuition into scalable rules is nontrivial. Multimodal LLMs (MLLMs) offer a promising solution, given their world knowledge, human-preference alignment, and reasoning skills. We evaluate MLLM verifiers across web navigation, computer use, and robotics, spanning 13 models, 28 judge designs, diverse agent trajectories, and distinct downstream applications. We identify a critical limitation: a strong tendency for MLLMs to over-validate agent behavior, a phenomenon we term agreement bias.