Act2Answer protocol reveals VLA models retain simple concepts but show larger gaps on complex semantics than source VLMs, with VQA co-training linked to better retention and knowledge signals peaking in middle layers.
Title resolution pending
12 Pith papers cite this work. Polarity classification is still indexing.
12
Pith papers citing it
citation-role summary
background 3
citation-polarity summary
years
2026 12roles
background 3polarities
background 3representative citing papers
GRA extracts 2D waypoints from synthetic videos to supervise VLA vision while restricting action training to real data, outperforming pseudo-action baselines on real-robot tasks.
FiberTune is a new fine-tuning objective that preserves action-fiber visual residuals in VLA policies, yielding performance gains on simulation and physical robot tasks.
LARA jointly optimizes LAM and VLA models via representation alignment to improve robotic manipulation performance using human videos.
AVP architecture has VLM emit visual-primitive tokens to condition flow-matching action expert, yielding 27.61% higher success rate than pi_0.5 on real-robot pick-and-place tasks.
VLA-AD distills 7B VLA teachers into 158M students using offline VLM semantic guidance on task phases and directions, matching teacher performance on LIBERO with 44x size reduction and 3.28x speedup.
Anchor-Centric Adaptation escapes the diversity trap by prioritizing repeated demonstrations at core anchors over broad coverage, yielding higher success rates under fixed data budgets in robotic manipulation.
DeLock mitigates lock-in in low-data VLA post-training via visual grounding preservation and test-time contrastive prompt guidance, outperforming baselines across eight evaluations while matching data-heavy generalist policies.
A vision-language-aligned world model turns visuomotor MPC into a language-following planner that reaches 87% success on 288 unseen semantic tasks where standard VLAs drop to 22%.
Pretraining a vision-language model to output discrete 3D pose tokens on large non-robotic data, before training a robot action head, improves downstream manipulation success and data efficiency.
Sharpness-aware minimization during VLA finetuning preserves instruction following and yields over 60% gains across simulation and real-world tasks.
EvoDriveVLA uses collaborative perception-planning distillation with self-anchor and future-aware teachers to fix perception degradation and long-term instability in driving VLA models, reaching SOTA on nuScenes and NAVSIM.
citing papers explorer
-
Does VLA Even Know the Basics? Measuring Commonsense and World Knowledge Retention in Vision-Language-Action Models
Act2Answer protocol reveals VLA models retain simple concepts but show larger gaps on complex semantics than source VLMs, with VQA co-training linked to better retention and knowledge signals peaking in middle layers.
-
Supervise What Survives: Geometry-Guided VLA Adaptation from Synthetic Robot Videos
GRA extracts 2D waypoints from synthetic videos to supervise VLA vision while restricting action training to real data, outperforming pseudo-action baselines on real-robot tasks.
-
FiberTune: Preserving Action-Fiber Visual Residuals in Vision-Language-Action Fine-Tuning
FiberTune is a new fine-tuning objective that preserves action-fiber visual residuals in VLA policies, yielding performance gains on simulation and physical robot tasks.
-
LARA: Latent Action Representation Alignment for Vision-Language-Action Models
LARA jointly optimizes LAM and VLA models via representation alignment to improve robotic manipulation performance using human videos.
-
Action with Visual Primitives
AVP architecture has VLM emit visual-primitive tokens to condition flow-matching action expert, yielding 27.61% higher success rate than pi_0.5 on real-robot pick-and-place tasks.
-
Offline Semantic Guidance for Efficient Vision-Language-Action Policy Distillation
VLA-AD distills 7B VLA teachers into 158M students using offline VLM semantic guidance on task phases and directions, matching teacher performance on LIBERO with 44x size reduction and 3.28x speedup.
-
Escaping the Diversity Trap in Robotic Manipulation via Anchor-Centric Adaptation
Anchor-Centric Adaptation escapes the diversity trap by prioritizing repeated demonstrations at core anchors over broad coverage, yielding higher success rates under fixed data budgets in robotic manipulation.
-
Breaking Lock-In: Preserving Steerability under Low-Data VLA Post-Training
DeLock mitigates lock-in in low-data VLA post-training via visual grounding preservation and test-time contrastive prompt guidance, outperforming baselines across eight evaluations while matching data-heavy generalist policies.
-
Grounded World Model for Semantically Generalizable Planning
A vision-language-aligned world model turns visuomotor MPC into a language-following planner that reaches 87% success on 288 unseen semantic tasks where standard VLAs drop to 22%.
-
PoseVLA: Universal Pose Pretraining for Generalizable Vision-Language-Action Policies
Pretraining a vision-language model to output discrete 3D pose tokens on large non-robotic data, before training a robot action head, improves downstream manipulation success and data efficiency.
-
Flatness Preserves Instruction Following in Vision-Language-Action Models
Sharpness-aware minimization during VLA finetuning preserves instruction following and yields over 60% gains across simulation and real-world tasks.
-
EvoDriveVLA: Evolving Driving VLA Models via Collaborative Perception-Planning Distillation
EvoDriveVLA uses collaborative perception-planning distillation with self-anchor and future-aware teachers to fix perception degradation and long-term instability in driving VLA models, reaching SOTA on nuScenes and NAVSIM.