arXiv subjects

Haojie Huang

Publications and source records attributed to Haojie Huang.

At least 19 recordsLinked to original sources

arXiv2026

Evidence-RL: Towards Evidence-intensive Visual Reasoning

Vision-Language Models (VLMs) should answer from concrete image evidence rather than language priors, dataset shortcuts, or irrelevant visual context. Existing perception-aware post-training methods encourage image use through global perturbations or attention proxies, but they do not test whether a sampled answer causally depends on the local evidence that supports it. We propose Counterfactual Evidence Disentanglement (CED), a training-time evidence audit for VLM grounding. For each response, CED neutralizes an object-centric Evidence Region and compares the resulting support drop against matched non-evidence Regions. We combine this signal with answer correctness inside GRPO, rewarding correct answers that rely on the evidence path rather than shortcut or nuisance paths. CED uses weak object-level proposals, requires no question-specific evidence annotations, and adds no inference-time overhead. Across nine public benchmarks and four backbones, CED outperforms prior RL-based post-training methods, with targeted analyses verifying its object-centric signal.

cs.CV↗

arXiv2026

Retriever: Composing the Perception-Reasoning-Action Loop for Long-Horizon Manipulation

Building long-horizon robot agents requires composing closed-loop pipelines -- perception, belief update, planning, and control -- whose components run at different clocks and with variable latency. Today, these systems are often assembled with ad-hoc concurrency and pub/sub conventions that make timing and input-consumption semantics implicit, yielding schedule-dependent behavior that is hard to reproduce, debug, and reuse. Current solutions typically solve parts of this problem at either the algorithmic or the systems layer, but not both. In this work, we propose Retriever, which spans the entire stack: an asynchronous decision model, a programming model, a runtime, and an example closed-loop agent pipeline. Retriever represents an agent as a graph of stateful causal stream functions executed on explicit run clocks. We formalize this view via an asynchronous environment-agent loop over continuous-time streams and show that finite-memory causal policies can be represented by compositions of these operators. Retriever compiles these graphs into a runtime that supports multiple backends, enabling systematic debugging across running environments and deterministic replay from logged asynchronous data. We evaluate Retriever through a real-robot case study together with controlled studies of runtime overhead and deterministic replay behavior.

cs.RO↗

arXiv2026

Dissecting Embodied Abilities in Multimodal Language Models through Skill-level Evaluation and Diagnosis

Understanding the capability bottlenecks of embodied multimodal large language models (MLLMs) is crucial for improving embodied agents. However, existing embodied benchmarks mainly focus on task-level evaluation and fail to provide actionable insights into the underlying causes of model failures. To address this limitation, we introduce BEAR, a benchmark that decomposes embodied tasks into 14 atomic skills for fine-grained skill-level evaluation. BEAR comprises 4,469 interleaved image-video-text samples spanning 14 skills across 6 categories, ranging from low-level perception to high-level planning. We evaluate 20 MLLMs on BEAR under a hierarchical skill-level diagnosis framework and uncover two key findings: (1) perceptual capabilities are major bottlenecks behind reasoning failures, and (2) current models suffer from unstable spatiotemporal modeling that remains largely unexposed in prior benchmarks. Motivated by these findings, we further propose BEAR-Agent, a multimodal conversational agent that augments MLLMs with visual and spatial reasoning tools. BEAR-Agent substantially improves performance across embodied skills, achieving a relative improvement of 17.5% on GPT-5 over the base model on BEAR, while also outperforming strong baselines in both simulation and real-world robotic experiments. Project page: https://bear-official66.github.io/

cs.CV↗

arXiv2026

Scale Up Strategically: Learning Compositional Generalization via Bias-Aware Evaluation and Data Collection for Robotic Manipulation

Compositional generalization is essential for robot to follow diverse instructions. However, pretrained policies are known to take shortcuts, deferring to salient cues rather than grounding language. We introduce a diagnostic framework that localizes this failure to individual \textit{instruction factors}, \textit{e.g.,} reusable semantic components such as color, verb, object, size, and spatial attribute. Our framework formalizes instruction factor bias, the tendency of fine-tuned policies to over-rely on dominant factors as shortcuts, and quantifies it through two metrics: Factor Dominance Rate (FDR), capturing pairwise bias between factors, and Factor Dominance Hierarchy (FDH), aggregating these into a global ranking. Evaluation on six foundation policies reveals broadly consistent ordering, \textit{i.e.}, color $\geq$ object $\geq$ spatial $\geq$ verb $\geq$ size, with color dominant, and verb and size most under-grounded. We further show the diagnosis is actionable: a bias-aware data collection strategy that reallocates a fixed budget toward under-grounded factors outperforms baselines in simulation and on a real robot using half the demonstrations, thereby enabling more sample-efficient and generalizable policy learning.

cs.RO↗

arXiv2026

Pix2Act: Image-Space Manipulation Policies with Equivariant Augmentation

Representing manipulation actions as 2D trajectories in the camera plane provides a compact and interpretable basis for learning complex 3D manipulation policies. However, it also creates challenges from out-of-frame trajectories and limited precision. We propose Pix2Act, an imitation learning method that addresses these challenges by generating continuous image-space keypoint trajectories in each camera plane and losslessly recovering end-effector poses via triangulation. This reformulates high-dimensional 3D control as a simpler, more learnable 2D prediction problem. Crucially, it aligns observations and actions in the same coordinate space, enabling equivariant transformations to jointly rotate individual camera images together with their image-space actions. We analyze the symmetry properties of this augmentation and design a network architecture that can fuse multiple camera views while respecting their per-view rotations. As a result, Pix2Act implicitly enlarges the support of the data distribution and learns invariant action structures across transformations, yielding improved generalization and overall performance. Across diverse simulated and real-world manipulation tasks, Pix2Act outperforms state-of-the-art baselines and remains robust under camera perturbations.

cs.RO↗

arXiv2026

Action Map Policy: Learning 3D Closed-loop Manipulation via Pixel Classification

The action space poses a major challenge in robot learning, since it is often high-dimensional, can span long time horizons, and frequently admits multi-modal optimal solutions. A good choice of action representation and loss function can help to address these concerns, but there are often trade offs. We propose Action Map Policy (AMP), which casts 3D closed-loop manipulation policy learning as a classification problem in image space. While classification has been an effective formulation in generative language models, applying it to robot action learning is difficult because naively discretizing high-dimensional continuous actions explodes the token vocabulary. Our key idea is to project 3D actions onto the camera image planes and treat each pixel location as a discrete class, thus controlling dimensionality while retaining multi-modality. This method supports millimeter-level precision for high-dimensional actions without requiring a prohibitively large vocabulary, while preserving fine-grained pixel-wise visual signals. Furthermore, it can predict the entire action chunk in a single forward pass, avoiding complex noise scheduling and iterative denoising while achieving substantially faster inference than diffusion policies. Experiments on various manipulation tasks show that AMP outperforms strong baselines, achieving higher success rates, faster inference, and enhanced spatial reasoning.

cs.RO↗

arXiv2026

The Latent Space: Foundation, Evolution, Mechanism, Ability, and Outlook

Latent space is rapidly emerging as a native substrate for language-based models. While modern systems are still commonly understood through explicit token-level generation, an increasing body of work shows that many critical internal processes are more naturally carried out in continuous latent space than in human-readable verbal traces. This shift is driven by the structural limitations of explicit-space computation, including linguistic redundancy, discretization bottlenecks, sequential inefficiency, and semantic loss. This survey aims to provide a unified and up-to-date landscape of latent space in language-based models. We organize the survey into five sequential perspectives: Foundation, Evolution, Mechanism, Ability, and Outlook. We begin by delineating the scope of latent space, distinguishing it from explicit or verbal space and from the latent spaces commonly studied in generative visual models. We then trace the field's evolution from early exploratory efforts to the current large-scale expansion. To organize the technical landscape, we examine existing work through the complementary lenses of mechanism and ability. From the perspective of Mechanism, we identify four major lines of development: Architecture, Representation, Computation, and Optimization. From the perspective of Ability, we show how latent space supports a broad capability spectrum spanning Reasoning, Planning, Modeling, Perception, Memory, Collaboration, and Embodiment. Beyond consolidation, we discuss the key open challenges, and outline promising directions for future research. We hope this survey serves not only as a reference for existing work, but also as a foundation for understanding latent space as a general computational and systems paradigm for next-generation intelligence.

cs.AI↗

arXiv2026

SandSim: Curve-Guided Gaussian Splatting for Reconstructing Sand Painting Processes

Sand painting is a process-driven art where visual appearance emerges from granular accumulation. Given a single image, reconstructing a plausible sand painting process requires modeling coherent stroke structures and material-dependent effects. Existing methods, including stroke-based optimization and diffusion-based video synthesis, often lack structural coherence and material consistency, leading to unrealistic drawing sequences. We present SandSim, a framework that reconstructs a sand painting process from a single image. We introduce a curve-guided Gaussian representation that models strokes as sequences of anisotropic primitives along continuous trajectories, whose smooth kernels capture the soft boundaries of sand strokes and enable coherent stroke formation. We further adopt a subtractive compositing scheme to model light attenuation during sand accumulation. We incorporate a semantic-guided planning module for scene decomposition and drawing order inference. Our framework jointly optimizes stroke geometry and appearance and can be integrated with a physics-based simulator for interactive sand dynamics and editing. Experiments show that our method produces temporally coherent and visually realistic results, achieving improved reconstruction quality and perceptual fidelity compared to existing approaches.

cs.GR↗

arXiv2026

ThinkGrasp: A Vision-Language System for Strategic Part Grasping in Clutter

Robotic grasping in cluttered environments remains a significant challenge due to occlusions and complex object arrangements. We have developed ThinkGrasp, a plug-and-play vision-language grasping system that makes use of GPT-4o's advanced contextual reasoning for heavy clutter environment grasping strategies. ThinkGrasp can effectively identify and generate grasp poses for target objects, even when they are heavily obstructed or nearly invisible, by using goal-oriented language to guide the removal of obstructing objects. This approach progressively uncovers the target object and ultimately grasps it with a few steps and a high success rate. In both simulated and real experiments, ThinkGrasp achieved a high success rate and significantly outperformed state-of-the-art methods in heavily cluttered environments or with diverse unseen objects, demonstrating strong generalization capabilities.

cs.RO↗

arXiv2026

MME-CoF-Pro: Evaluating Reasoning Coherence in Video Generative Models with Text and Visual Hints

Video generative models show emerging reasoning behaviors. It is essential to ensure that generated events remain causally consistent across frames for reliable deployment, a property we define as reasoning coherence. To bridge the gap in literature for missing reasoning coherence evaluation, we propose MME-CoF-Pro, a comprehensive video reasoning benchmark to assess reasoning coherence in video models. Specifically, MME-CoF-Pro contains 303 samples across 16 categories, ranging from visual logical to scientific reasoning. It introduces Reasoning Score as evaluation metric for assessing process-level necessary intermediate reasoning steps, and includes three evaluation settings, (a) no hint (b) text hint and (c) visual hint, enabling a controlled investigation into the underlying mechanisms of reasoning hint guidance. Evaluation results in 7 open and closed-source video models reveals insights including: (1) Video generative models exhibit weak reasoning coherence, decoupled from generation quality. (2) Text hints boost apparent correctness but often cause inconsistency and hallucinated reasoning (3) Visual hints benefit structured perceptual tasks but struggle with fine-grained perception. Website: https://video-reasoning-coherence.github.io/

cs.CV↗

arXiv2025

3D Equivariant Visuomotor Policy Learning via Spherical Projection

Equivariant models have recently been shown to improve the data efficiency of diffusion policy by a significant margin. However, prior work that explored this direction focused primarily on point cloud inputs generated by multiple cameras fixed in the workspace. This type of point cloud input is not compatible with the now-common setting where the primary input modality is an eye-in-hand RGB camera like a GoPro. This paper closes this gap by incorporating into the diffusion policy model a process that projects features from the 2D RGB camera image onto a sphere. This enables us to reason about symmetries in $\mathrm{SO}(3)$ without explicitly reconstructing a point cloud. We perform extensive experiments in both simulation and the real world that demonstrate that our method consistently outperforms strong baselines in terms of both performance and sample efficiency. Our work, Image-to-Sphere Policy ($\textbf{ISP}$), is the first $\mathrm{SO}(3)$-equivariant policy learning framework for robotic manipulation that works using only monocular RGB inputs.

cs.RO↗

arXiv2025

Generalizable Hierarchical Skill Learning via Object-Centric Representation

We present Generalizable Hierarchical Skill Learning (GSL), a novel framework for hierarchical policy learning that significantly improves policy generalization and sample efficiency in robot manipulation. One core idea of GSL is to use object-centric skills as an interface that bridges the high-level vision-language model and the low-level visual-motor policy. Specifically, GSL decomposes demonstrations into transferable and object-canonicalized skill primitives using foundation models, ensuring efficient low-level skill learning in the object frame. At test time, the skill-object pairs predicted by the high-level agent are fed to the low-level module, where the inferred canonical actions are mapped back to the world frame for execution. This structured yet flexible design leads to substantial improvements in sample efficiency and generalization of our method across unseen spatial arrangements, object appearances, and task compositions. In simulation, GSL trained with only 3 demonstrations per task outperforms baselines trained with 30 times more data by 15.5 percent on unseen tasks. In real-world experiments, GSL also surpasses the baseline trained with 10 times more data.

cs.RO↗

arXiv2025

Clebsch-Gordan Transformer: Fast and Global Equivariant Attention

The global attention mechanism is one of the keys to the success of transformer architecture, but it incurs quadratic computational costs in relation to the number of tokens. On the other hand, equivariant models, which leverage the underlying geometric structures of problem instance, often achieve superior accuracy in physical, biochemical, computer vision, and robotic tasks, at the cost of additional compute requirements. As a result, existing equivariant transformers only support low-order equivariant features and local context windows, limiting their expressiveness and performance. This work proposes Clebsch-Gordan Transformer, achieving efficient global attention by a novel Clebsch-Gordon Convolution on $\SO(3)$ irreducible representations. Our method enables equivariant modeling of features at all orders while achieving ${O}(N \log N)$ input token complexity. Additionally, the proposed method scales well with high-order irreducible features, by exploiting the sparsity of the Clebsch-Gordon matrix. Lastly, we also incorporate optional token permutation equivariance through either weight sharing or data augmentation. We benchmark our method on a diverse set of benchmarks including n-body simulation, QM9, ModelNet point cloud classification and a robotic grasping dataset, showing clear gains over existing equivariant transformers in GPU memory size, speed, and accuracy.

cs.LG↗

arXiv2025

Equivariant Goal Conditioned Contrastive Reinforcement Learning

Contrastive Reinforcement Learning (CRL) provides a promising framework for extracting useful structured representations from unlabeled interactions. By pulling together state-action pairs and their corresponding future states, while pushing apart negative pairs, CRL enables learning nontrivial policies without manually designed rewards. In this work, we propose Equivariant CRL (ECRL), which further structures the latent space using equivariant constraints. By leveraging inherent symmetries in goal-conditioned manipulation tasks, our method improves both sample efficiency and spatial generalization. Specifically, we formally define Goal-Conditioned Group-Invariant MDPs to characterize rotation-symmetric robotic manipulation tasks, and build on this by introducing a novel rotation-invariant critic representation paired with a rotation-equivariant actor for Contrastive RL. Our approach consistently outperforms strong baselines across a range of simulated tasks in both state-based and image-based settings. Finally, we extend our method to the offline RL setting, demonstrating its effectiveness across multiple tasks.

cs.RO↗

arXiv2025

Learning Efficient and Robust Language-conditioned Manipulation using Textual-Visual Relevancy and Equivariant Language Mapping

Controlling robots through natural language is pivotal for enhancing human-robot collaboration and synthesizing complex robot behaviors. Recent works that are trained on large robot datasets show impressive generalization abilities. However, such pretrained methods are (1) often fragile to unseen scenarios, and (2) expensive to adapt to new tasks. This paper introduces Grounded Equivariant Manipulation (GEM), a robust yet efficient approach that leverages pretrained vision-language models with equivariant language mapping for language-conditioned manipulation tasks. Our experiments demonstrate GEM's high sample efficiency and generalization ability across diverse tasks in both simulation and the real world. GEM achieves similar or higher performance with orders of magnitude fewer robot data compared with major data-efficient baselines such as CLIPort and VIMA. Finally, our approach demonstrates greater robustness compared to large VLA model, e.g, OpenVLA, at correctly interpreting natural language commands on unseen objects and poses. Code, data, and training details are available https://saulbatman.github.io/gem_page/

cs.RO↗

arXiv2025

Push-Grasp Policy Learning Using Equivariant Models and Grasp Score Optimization

Goal-conditioned robotic grasping in cluttered environments remains a challenging problem due to occlusions caused by surrounding objects, which prevent direct access to the target object. A promising solution to mitigate this issue is combining pushing and grasping policies, enabling active rearrangement of the scene to facilitate target retrieval. However, existing methods often overlook the rich geometric structures inherent in such tasks, thus limiting their effectiveness in complex, heavily cluttered scenarios. To address this, we propose the Equivariant Push-Grasp Network, a novel framework for joint pushing and grasping policy learning. Our contributions are twofold: (1) leveraging SE(2)-equivariance to improve both pushing and grasping performance and (2) a grasp score optimization-based training strategy that simplifies the joint learning process. Experimental results show that our method improves grasp success rates by 49% in simulation and by 35% in real-world scenarios compared to strong baselines, representing a significant advancement in push-grasp policy learning.

cs.RO↗

arXiv2025

MATCH POLICY: A Simple Pipeline from Point Cloud Registration to Manipulation Policies

Many manipulation tasks require the robot to rearrange objects relative to one another. Such tasks can be described as a sequence of relative poses between parts of a set of rigid bodies. In this work, we propose MATCH POLICY, a simple but novel pipeline for solving high-precision pick and place tasks. Instead of predicting actions directly, our method registers the pick and place targets to the stored demonstrations. This transfers action inference into a point cloud registration task and enables us to realize nontrivial manipulation policies without any training. MATCH POLICY is designed to solve high-precision tasks with a key-frame setting. By leveraging the geometric interaction and the symmetries of the task, it achieves extremely high sample efficiency and generalizability to unseen configurations. We demonstrate its state-of-the-art performance across various tasks on RLBench benchmark compared with several strong baselines and test it on a real robot with six tasks.

cs.RO↗

arXiv2025

Coarse-to-Fine 3D Keyframe Transporter

Recent advances in Keyframe Imitation Learning (IL) have enabled learning-based agents to solve a diverse range of manipulation tasks. However, most approaches ignore the rich symmetries in the problem setting and, as a consequence, are sample-inefficient. This work identifies and utilizes the bi-equivariant symmetry within Keyframe IL to design a policy that generalizes to transformations of both the workspace and the objects grasped by the gripper. We make two main contributions: First, we analyze the bi-equivariance properties of the keyframe action scheme and propose a Keyframe Transporter derived from the Transporter Networks, which evaluates actions using cross-correlation between the features of the grasped object and the features of the scene. Second, we propose a computationally efficient coarse-to-fine SE(3) action evaluation scheme for reasoning the intertwined translation and rotation action. The resulting method outperforms strong Keyframe IL baselines by an average of >10% on a wide range of simulation tasks, and by an average of 55% in 4 physical experiments.

cs.RO↗