Hung Luong Viet

Vision-language models (VLMs) have achieved strong performance across diverse multimodal tasks, yet they remain vulnerable to unreliable reasoning. Existing self-correction methods mitigate these issues but typically rely on post-training or carefully engineered feedback, incurring high computational cost. In this work, we revisit this challenge through the lens of emotional cues, asking whether they can activate latent self-correction behaviors in VLMs without additional training.

Article Dans Une Revue Conservation Letters Année : 2024 Addressing the Southeast Asian snaring crisis: Impact of 11 years of snare removal in a biodiversity hotspot Mots clés snaring Southeast Asia Viet Nam biodiversity conservation protected area Domaines Biodiversité et Ecologie Ecosystèmes Génétique des populations [q-bio.PE] Fichier principal Conservation Letters - 2024 - Tilker - Addressing the Southeast Asian snaring crisis Impact of 11 years of snare removal.pdf (927.22 Ko)...

1 INTRODUCTION Unsustainable hunting has defaunated tropical forests around the world (Benítez-López et al., 2017, 2019; Harrison et al., 2016), and preventing further defaunation is a priority for addressing global biodiversity loss (Barlow et al., 2018; Laurance et al., 2012).