5 citations · 6 across the 9 of their papers we have counts for
Showing 2026Show all
3 papers · 1 filter
cs.CV2026
Cosmos 3: Omnimodal World Models for Physical AI
NVIDIA, :, Aditi +293
We introduce Cosmos 3, a family of omnimodal world models designed to jointly process and generate language, image, video, audio, and action sequences within a unified mixture-of-t…
cs.CV2026
Beyond Encoder Accumulation: Measuring Encoder Roles in Multi-Encoder VLMs
Wei Ding, Yudong Zhang, Ruobing Xie +3
As foundation models scale toward fusing more heterogeneous visual streams, understanding how diverse encoders interact under joint training becomes a prerequisite for principled d…
cs.CV2026
MHSA: A Lightweight Framework for Mitigating Hallucinations via Steered Attention in LVLMs
Wei Ding, Yilin Li, Yudong Zhang +4
Large vision-language models (LVLMs) have achieved remarkable performance across diverse multimodal tasks, yet they continue to suffer from hallucinations, generating content that…