VEGA improves spatial reasoning in VLA models for robotics by aligning visual encoder features with 3D-supervised DINOv2 representations via a temporary projector and cosine similarity loss.
arXiv preprint arXiv:2512.09619 (2025)
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
baseline 1
citation-polarity summary
years
2026 3verdicts
UNVERDICTED 3representative citing papers
Distillation of a 688M-parameter MASt3R teacher yields up to 7x smaller students that retain most lunar reconstruction accuracy and outperform sparse-supervised baselines.
X-Imitator is a bidirectional action-pose interaction framework for spatial-aware imitation learning that outperforms vanilla policies and explicit pose guidance on 24 simulated and 3 real-world robotic tasks.
citing papers explorer
-
VEGA: Visual Encoder Grounding Alignment for Spatially-Aware Vision-Language-Action Models
VEGA improves spatial reasoning in VLA models for robotics by aligning visual encoder features with 3D-supervised DINOv2 representations via a temporary projector and cosine similarity loss.
-
Geometric Foundation Model Distillation for Efficient Lunar 3D Reconstruction
Distillation of a 688M-parameter MASt3R teacher yields up to 7x smaller students that retain most lunar reconstruction accuracy and outperform sparse-supervised baselines.
-
X-Imitator: Spatial-Aware Imitation Learning via Bidirectional Action-Pose Interaction
X-Imitator is a bidirectional action-pose interaction framework for spatial-aware imitation learning that outperforms vanilla policies and explicit pose guidance on 24 simulated and 3 real-world robotic tasks.