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.
hub Canonical reference
An Embodied Generalist Agent in 3D World
Canonical reference. 100% of citing Pith papers cite this work as background.
25
Pith papers citing it
Background
100% of classified citations
abstract
Leveraging massive knowledge from large language models (LLMs), recent machine learning models show notable successes in general-purpose task solving in diverse domains such as computer vision and robotics. However, several significant challenges remain: (i) most of these models rely on 2D images yet exhibit a limited capacity for 3D input; (ii) these models rarely explore the tasks inherently defined in 3D world, e.g., 3D grounding, embodied reasoning and acting. We argue these limitations significantly hinder current models from performing real-world tasks and approaching general intelligence. To this end, we introduce LEO, an embodied multi-modal generalist agent that excels in perceiving, grounding, reasoning, planning, and acting in the 3D world. LEO is trained with a unified task interface, model architecture, and objective in two stages: (i) 3D vision-language (VL) alignment and (ii) 3D vision-language-action (VLA) instruction tuning. We collect large-scale datasets comprising diverse object-level and scene-level tasks, which require considerable understanding of and interaction with the 3D world. Moreover, we meticulously design an LLM-assisted pipeline to produce high-quality 3D VL data. Through extensive experiments, we demonstrate LEO's remarkable proficiency across a wide spectrum of tasks, including 3D captioning, question answering, embodied reasoning, navigation and manipulation. Our ablative studies and scaling analyses further provide valuable insights for developing future embodied generalist agents. Code and data are available on project page.
hub tools
citation-role summary
background 6
citation-polarity summary
roles
background 6polarities
background 6representative citing papers
PokeGym is a new benchmark that tests VLMs on long-horizon tasks in a complex 3D game using only visual observations, identifying deadlock recovery as the primary failure mode.
POMA-3D learns self-supervised 3D scene representations from point maps and improves performance on geometric 3D tasks including navigation and scene retrieval.
REI-Bench shows vague referring expressions degrade LLM robot planning success by up to 36.9%, with task-oriented context cognition providing effective mitigation.
3D-VLA is a new embodied foundation model that uses a 3D LLM plus aligned diffusion models to generate future images and point clouds for improved reasoning and action planning in 3D environments.
DepthVLM converts a standard VLM into a dense metric depth predictor by attaching a lightweight head and training under unified vision-text supervision, outperforming prior VLMs and some pure vision models on a new indoor-outdoor benchmark.
ConsisVLA-4D adds cross-view semantic alignment, cross-object geometric fusion, and cross-scene dynamic reasoning to VLA models, delivering 21.6% and 41.5% gains plus 2.3x and 2.4x speedups on LIBERO and real-world tasks.
Unlabeled web videos processed by designed data engines generate effective training data that yields strong zero-shot and finetuned performance on 3D detection, segmentation, VQA, and navigation.
Chat-Scene++ improves 3D scene understanding in multimodal LLMs by representing scenes as context-rich object sequences with identifier tokens and grounded chain-of-thought reasoning, reaching state-of-the-art on five benchmarks using pre-trained encoders.
SpatialStack improves 3D spatial reasoning in vision-language models by stacking and synchronizing multi-level geometric features with the language backbone.
Motion-MLLM integrates IMU egomotion data into MLLMs using cascaded filtering and asymmetric fusion to ground visual content in physical trajectories for scale-aware 3D understanding, achieving competitive accuracy at higher speed.
TrianguLang achieves state-of-the-art feed-forward text-guided 3D localization and segmentation by using predicted geometry to gate cross-view semantic correspondences without ground-truth poses.
OptimusVLA augments hierarchical VLA models with Global Prior Memory for shorter generative paths and Local Consistency Memory for temporal coherence, yielding higher success rates and 2.9x faster inference on simulation and real-world robotic benchmarks.
CLAMP pretrains 3D multi-view encoders with contrastive learning on point clouds and actions, then initializes diffusion policies for more sample-efficient fine-tuning on robotic tasks.
EyeVLA transfers open-world VLM understanding to a PTZ camera control policy via hierarchical action tokens and GRPO reinforcement learning, reaching 96% task completion on 50 real scenes with only 500 training samples.
C-Nav is a continual visual navigation framework with dual-path anti-forgetting via feature distillation and replay plus adaptive sampling that outperforms baselines on a new continual object navigation benchmark while using less memory.
DEGround presents a unified homogeneous framework for 3D visual grounding with shared queries and two plug-in modules for better instruction alignment, reporting a 7.52% improvement on the EmbodiedScan benchmark.
Spatial-MLLM adds a 3D spatial encoder initialized from a visual geometry model and space-aware frame sampling to MLLMs to improve spatial understanding and reasoning from purely 2D visual inputs.
A principled reward design for tool selection and application in RL-trained LLMs delivers 17% gains over base models and 15% over SFT across benchmarks.
ValuePlanner is a hierarchical architecture that uses LLMs to generate value-based subgoals and PDDL planners to produce executable actions, enabling self-directed behavior in embodied agents.
The survey frames VLA models as pipelines that generate progressively grounded action tokens and classifies those tokens into eight types to guide future development.
WorldVLA unifies VLA and world models in one autoregressive system, shows they boost each other, and adds an attention mask to stop error buildup when generating action chunks.
MLLMs achieve only 42% accuracy on a new audio-visual task requiring second-order spatial ToM under perceptual limits, while a proposed sensory-bounded CoT outperforms egocentric and allocentric baselines.
AugVLA-3D augments existing VLA models with depth-derived 3D features and action priors to improve generalization and action accuracy in 3D 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.
-
PokeGym: A Visually-Driven Long-Horizon Benchmark for Vision-Language Models
PokeGym is a new benchmark that tests VLMs on long-horizon tasks in a complex 3D game using only visual observations, identifying deadlock recovery as the primary failure mode.
-
POMA-3D: The Point Map Way to 3D Scene Understanding
POMA-3D learns self-supervised 3D scene representations from point maps and improves performance on geometric 3D tasks including navigation and scene retrieval.
-
REI-Bench: Can Embodied Agents Understand Vague Human Instructions in Task Planning?
REI-Bench shows vague referring expressions degrade LLM robot planning success by up to 36.9%, with task-oriented context cognition providing effective mitigation.
-
3D-VLA: A 3D Vision-Language-Action Generative World Model
3D-VLA is a new embodied foundation model that uses a 3D LLM plus aligned diffusion models to generate future images and point clouds for improved reasoning and action planning in 3D environments.
-
Unlocking Dense Metric Depth Estimation in VLMs
DepthVLM converts a standard VLM into a dense metric depth predictor by attaching a lightweight head and training under unified vision-text supervision, outperforming prior VLMs and some pure vision models on a new indoor-outdoor benchmark.
-
ConsisVLA-4D: Advancing Spatiotemporal Consistency in Efficient 3D-Perception and 4D-Reasoning for Robotic Manipulation
ConsisVLA-4D adds cross-view semantic alignment, cross-object geometric fusion, and cross-scene dynamic reasoning to VLA models, delivering 21.6% and 41.5% gains plus 2.3x and 2.4x speedups on LIBERO and real-world tasks.
-
Lifting Unlabeled Internet-level Data for 3D Scene Understanding
Unlabeled web videos processed by designed data engines generate effective training data that yields strong zero-shot and finetuned performance on 3D detection, segmentation, VQA, and navigation.
-
Chat-Scene++: Exploiting Context-Rich Object Identification for 3D LLM
Chat-Scene++ improves 3D scene understanding in multimodal LLMs by representing scenes as context-rich object sequences with identifier tokens and grounded chain-of-thought reasoning, reaching state-of-the-art on five benchmarks using pre-trained encoders.
-
SpatialStack: Layered Geometry-Language Fusion for 3D VLM Spatial Reasoning
SpatialStack improves 3D spatial reasoning in vision-language models by stacking and synchronizing multi-level geometric features with the language backbone.
-
Feeling the Space: Egomotion-Aware Video Representation for Efficient and Accurate 3D Scene Understanding
Motion-MLLM integrates IMU egomotion data into MLLMs using cascaded filtering and asymmetric fusion to ground visual content in physical trajectories for scale-aware 3D understanding, achieving competitive accuracy at higher speed.
-
TrianguLang: Geometry-Aware Semantic Consensus for Pose-Free 3D Localization
TrianguLang achieves state-of-the-art feed-forward text-guided 3D localization and segmentation by using predicted geometry to gate cross-view semantic correspondences without ground-truth poses.
-
Global Prior Meets Local Consistency: Dual-Memory Augmented Vision-Language-Action Model for Efficient Robotic Manipulation
OptimusVLA augments hierarchical VLA models with Global Prior Memory for shorter generative paths and Local Consistency Memory for temporal coherence, yielding higher success rates and 2.9x faster inference on simulation and real-world robotic benchmarks.
-
CLAMP: Contrastive Learning for 3D Multi-View Action-Conditioned Robotic Manipulation Pretraining
CLAMP pretrains 3D multi-view encoders with contrastive learning on point clouds and actions, then initializes diffusion policies for more sample-efficient fine-tuning on robotic tasks.
-
Look, Zoom, Understand: The Robotic Eyeball for Embodied Perception
EyeVLA transfers open-world VLM understanding to a PTZ camera control policy via hierarchical action tokens and GRPO reinforcement learning, reaching 96% task completion on 50 real scenes with only 500 training samples.
-
C-NAV: Towards Self-Evolving Continual Object Navigation in Open World
C-Nav is a continual visual navigation framework with dual-path anti-forgetting via feature distillation and replay plus adaptive sampling that outperforms baselines on a new continual object navigation benchmark while using less memory.
-
DEGround: An Effective Baseline for Ego-centric 3D Visual Grounding with a Homogeneous Framework
DEGround presents a unified homogeneous framework for 3D visual grounding with shared queries and two plug-in modules for better instruction alignment, reporting a 7.52% improvement on the EmbodiedScan benchmark.
-
Spatial-MLLM: Boosting MLLM Capabilities in Visual-based Spatial Intelligence
Spatial-MLLM adds a 3D spatial encoder initialized from a visual geometry model and space-aware frame sampling to MLLMs to improve spatial understanding and reasoning from purely 2D visual inputs.
-
ToolRL: Reward is All Tool Learning Needs
A principled reward design for tool selection and application in RL-trained LLMs delivers 17% gains over base models and 15% over SFT across benchmarks.
-
Bridging Values and Behavior: A Hierarchical Framework for Proactive Embodied Agents
ValuePlanner is a hierarchical architecture that uses LLMs to generate value-based subgoals and PDDL planners to produce executable actions, enabling self-directed behavior in embodied agents.
-
A Survey on Vision-Language-Action Models: An Action Tokenization Perspective
The survey frames VLA models as pipelines that generate progressively grounded action tokens and classifies those tokens into eight types to guide future development.
-
WorldVLA: Towards Autoregressive Action World Model
WorldVLA unifies VLA and world models in one autoregressive system, shows they boost each other, and adds an attention mask to stop error buildup when generating action chunks.
-
Beyond the Cartesian Illusion: Testing Two-Stage Multi-Modal Theory of Mind under Perceptual Bottlenecks
MLLMs achieve only 42% accuracy on a new audio-visual task requiring second-order spatial ToM under perceptual limits, while a proposed sensory-bounded CoT outperforms egocentric and allocentric baselines.
-
AugVLA-3D: Depth-Driven Feature Augmentation for Vision-Language-Action Models
AugVLA-3D augments existing VLA models with depth-derived 3D features and action priors to improve generalization and action accuracy in 3D robotic tasks.
-
A Survey on Multimodal Large Language Models
This survey organizes the architectures, training strategies, data, evaluation methods, extensions, and challenges of Multimodal Large Language Models.