Searcharxiv⌕ Search

arXiv subjects

Jing Huo

Publications and source records attributed to Jing Huo.

At least 19 recordsLinked to original sources
arXiv2026

When Teacher Guidance Misleads: Reward-Aligned On-Policy Distillation

On-policy distillation (OPD) has recently emerged as a popular post-training paradigm for large language models (LLMs), providing an efficient way to transfer the knowledge and capabilities of teacher models into student models. However, teacher guidance on student-generated prefixes is not always reliable. Training should optimize the model to generate responses that are more likely to be correct, or equivalently, to get higher outcome rewards. But during OPD, the teacher model may provide guidance that discourages the student from moving toward correct trajectories or moves the student toward incorrect ones, which is misaligned with outcome reward. Such misaligned guidance is unreliable, as it would mislead the optimization process and ultimately degrade model performance. To mitigate misaligned teacher guidance, we propose Reward-Aligned On-Policy Distillation (RA-OPD). The key insight is to keep only trajectories whose induced updates move the student toward correct trajectories or discourage the student from moving toward incorrect ones. Specifically, for each sampled trajectory, RA-OPD checks whether its trajectory-level distillation return is consistent with its outcome reward and then filters out the misaligned trajectories. RA-OPD selects more reliable trajectories to improve student model performance without requiring additional computational cost. We evaluate RA-OPD on math and code benchmarks using models from the Qwen3 family and the DeepSeek-R1 family. Across seven math benchmarks and three code benchmarks, RA-OPD significantly outperforms standard OPD and other tested OPD variants.

cs.AI↗
arXiv2026

Learning from Hard Prompts: Difficulty-aware Advantage Amplification in Dynamic Sampling

Decoupled Clip and Dynamic Sampling Policy Optimization (DAPO) is a prominent variant of Group Relative Policy Optimization (GRPO). DAPO introduces several improvements over GRPO. Among these, Dynamic Sampling contributes the most to DAPO's accuracy gains relative to GRPO. To improve accuracy, Dynamic Sampling enhances training stability by eliminating zero policy gradients from zero advantages. Specifically, it avoids such zero gradients by filtering out prompts where sampled responses are either entirely correct or incorrect. However, our theoretical analysis shows that Dynamic Sampling decrease training efficiency as it cannot effectively utilize hard-to-sample correct responses on hard prompts. Formally, it asymmetrically amplifies the advantages of distinct responses to the same prompts. On hard prompts, incorrect responses undergo greater amplification than correct ones. This leads the model to avoid generating the observed incorrect responses rather than capitalizing on the hard-to-sample correct ones on hard prompts, resulting in low training efficiency. To improve training efficiency, we propose Direct Advantage Amplification (DAA), which amplifies the advantages of hard-to-sample correct responses on hard prompts, as obtained by Dynamic Sampling. This ensures that, when Dynamic Sampling is used, these hard-to-sample responses can be effectively capitalized on, implying higher training efficiency. By integrating DAA into DAPO, we obtain Difficulty-aware Advantage Amplification Policy Optimization (DA3PO), which is implemented with fewer than 30 lines of code from DAPO. Experiments show that DA3PO significantly outperforms GRPO and other classical GRPO variants.

cs.AI↗
arXiv2026

Wh0: Generative World Models as Scalable Sources of Egocentric Human Hand Manipulation Data

Scaling dexterous manipulation requires generalization across objects, scenes, and tasks, yet existing data sources face a trade-off between scale and scene/embodiment alignment: teleoperation data is well aligned with robot deployment but expensive to collect; simulation is scalable but limited by the sim-to-real gap; and real egocentric videos scale effectively but remain misaligned with robot deployment. We propose Wh0, a framework that uses generative video world models as scalable and controllable sources of egocentric human-hand manipulation data to unlock the manipulation capabilities of pretrained dexterous VLA models. Conditioned on language, objects, and scenes, Wh0 uses a generative world model to produce WM-H, a 50k-episode dataset of egocentric human-object interaction videos. Wh0 then converts the generated videos into robot-trainable supervision through hand motion reconstruction and visual editing. Co-trained with a limited amount of real robot data, WM-H adapts pretrained VLA models to dexterous manipulation deployment. Across 18 real-world dexterous manipulation tasks, compared with a model post-trained only on robot data, Wh0 improves zero-shot success on unseen tasks from 8.3% to 38.9%. Ablation studies further show that scalable generation and scene/embodiment alignment are key drivers of performance gains. Videos and open-source code can be found on our project website: https://chenyt31.github.io/wh0.github.io/.

cs.RO↗
arXiv2026

Coarse-to-Control: Action-Token Planning for Vision-Language-Action Models

Most vision-language-action (VLA) models map observations directly to actions without explicit intermediate planning, which limits performance on long-horizon tasks where early mistakes compound. We propose Coarse-to-Control, a plan-execute VLA that introduces planning natively in the action-token space. The key idea is to let the policy first predict a compact sequence of coarse action tokens that summarize the intended future trajectory, and then generate executable action tokens conditioned on this plan. Because both planning and execution share a unified discrete action vocabulary, the plan stays close to the control manifold and provides directly actionable guidance rather than an abstract hint that must be translated back to motor commands. Experiments on LIBERO, SimplerEnv-WidowX, and real-world manipulation tasks show that action-token planning consistently improves over direct action generation, with the largest gains on long-horizon multi-stage tasks.

cs.RO↗
arXiv2026

Tool-Aware Optimization with Entropy Guidance for Efficient Agentic Reinforcement Learning

Agentic reinforcement learning (RL) equips large language models (LLMs) with tool-use capabilities that substantially improve reasoning on complex tasks. However, integrating external tools often destabilizes training: over-reliance on tools can induce input distribution shift, while overly conservative tool use limits effective exploration. To address this issue, we propose a unified framework TAO-RL that couples tool-aware trajectory filtering with entropy-guided exploration for efficient policy optimization. Specifically, at the data level, TAO-RL filters rollout trajectories along two criteria: discarding those where all tool invocations fail to execute, and removing those where all rollouts are either correct or incorrect, as both cases yield degenerate advantage estimates that contribute no discriminative learning signal. This joint filtering retains data that are both tool-capable and informative, establishing a high-quality training distribution. At the algorithmic level, we introduce a tool-aware entropy-guided bonus that reshapes the advantage function at post-tool-call tokens, encouraging the policy to explore more diverse reasoning paths at critical decision points. These two components are mutually reinforcing: trajectory filtering establishes a clean and informative training foundation, while entropy-guided exploration drives stronger reasoning behaviors at critical tool-interaction junctures. Extensive experiments on 7 challenging reasoning benchmarks across 3 model scales demonstrate the superiority of TAO-RL over existing methods.

cs.LG↗
arXiv2026

Uni-LaViRA: Language-Vision-Robot Actions Translation for Unified Embodied Navigation

Embodied navigation requires an agent to map language and visual observations to a stream of spatial actions that drive a real robot through environments it has never seen. The dominant approach has been to scale vision-language-action (VLA) foundation models on ever-larger collections of robot trajectories. This paper argues that, for navigation specifically, generality can be obtained structurally, not only through data scale. The underlying decision structure of navigation reduces to a single Language-Vision-Robot Actions Translation. The language action emits semantic-level directional command and the vision action emits a pixel-level visual target. Both outputs lie inside the natural output manifold of pretrained multimodal large language models (MLLMs), so the task can be reasoned about by an agent rather than learned from robot data. Therefore, we present Uni-LaViRA, a unified agentic architecture that extends the same insight to four task families (VLN-CE, ObjectNav, EQA, and Aerial-VLN) and to four heterogeneous real robots (Wheeled, Quadruped, Humanoid robot, and a self-built UAV) in a zero-shot manner. Two agent-loop mechanisms make this unification practical. TODO List Memory (TDM) rewrites a structured checklist of pending sub-goals at every step, reciting the unfinished items back into the agent's most recent attention window. Second Chance Backtrack (SCB) rolls the robot back to the pre-error state and conditions the agent's next plan on the failed sub-trajectory, turning single-pass navigation into a self-correcting process. With zero training effort, Uni-LaViRA reaches 60.7% SR on VLN-CE R2R, 51.3% on VLN-CE RxR, 77.7% on HM3D-v2, 60.0% on HM3D-OVON, 54.7% on MP3D-EQA, and 40.0% on OpenUAV, matching or even surpassing recent training navigation foundation models that consume millions of samples and thousands of GPU-hours.

cs.RO↗
arXiv2026

FineCog-Nav: Integrating Fine-grained Cognitive Modules for Zero-shot Multimodal UAV Navigation

UAV vision-language navigation (VLN) requires an agent to navigate complex 3D environments from an egocentric perspective while following ambiguous multi-step instructions over long horizons. Existing zero-shot methods remain limited, as they often rely on large base models, generic prompts, and loosely coordinated modules. In this work, we propose FineCog-Nav, a top-down framework inspired by human cognition that organizes navigation into fine-grained modules for language processing, perception, attention, memory, imagination, reasoning, and decision-making. Each module is driven by a moderate-sized foundation model with role-specific prompts and structured input-output protocols, enabling effective collaboration and improved interpretability. To support fine-grained evaluation, we construct AerialVLN-Fine, a curated benchmark of 300 trajectories derived from AerialVLN, with sentence-level instruction-trajectory alignment and refined instructions containing explicit visual endpoints and landmark references. Experiments show that FineCog-Nav consistently outperforms zero-shot baselines in instruction adherence, long-horizon planning, and generalization to unseen environments. These results suggest the effectiveness of fine-grained cognitive modularization for zero-shot aerial navigation. Project page: https://smartdianlab.github.io/projects-FineCogNav.

cs.CV↗
arXiv2026

V-Dreamer: Automating Robotic Simulation and Trajectory Synthesis via Video Generation Priors

Training generalist robots demands large-scale, diverse manipulation data, yet real-world collection is prohibitively expensive, and existing simulators are often constrained by fixed asset libraries and manual heuristics. To bridge this gap, we present V-Dreamer, a fully automated framework that generates open-vocabulary, simulation-ready manipulation environments and executable expert trajectories directly from natural language instructions. V-Dreamer employs a novel generative pipeline that constructs physically grounded 3D scenes using large language models and 3D generative models, validated by geometric constraints to ensure stable, collision-free layouts. Crucially, for behavior synthesis, we leverage video generation models as rich motion priors. These visual predictions are then mapped into executable robot trajectories via a robust Sim-to-Gen visual-kinematic alignment module utilizing CoTracker3 and VGGT. This pipeline supports high visual diversity and physical fidelity without manual intervention. To evaluate the generated data, we train imitation learning policies on synthesized trajectories encompassing diverse object and environment variations. Extensive evaluations on tabletop manipulation tasks using the Piper robotic arm demonstrate that our policies robustly generalize to unseen objects in simulation and achieve effective sim-to-real transfer, successfully manipulating novel real-world objects.

cs.RO↗
arXiv2026

ST-VLA: Enabling 4D-Aware Spatiotemporal Understanding for General Robot Manipulation

Robotic manipulation in open-world environments requires reasoning across semantics, geometry, and long-horizon action dynamics. Existing hierarchical Vision-Language-Action (VLA) frameworks typically use 2D representations to connect high-level reasoning with low-level control, but lack depth awareness and temporal consistency, limiting robustness in complex 3D scenes. We propose ST-VLA, a hierarchical VLA framework using a unified 3D-4D representation to bridge perception and action. ST-VLA converts 2D guidance into 3D trajectories and generates smooth spatial masks that capture 4D spatio-temporal context, providing a stable interface between semantic reasoning and continuous control. To enable effective learning of such representations, we introduce ST-Human, a large-scale human manipulation dataset with 14 tasks and 300k episodes, annotated with 2D, 3D, and 4D supervision via a semi-automated pipeline. Using ST-Human, we train ST-VLM, a spatio-temporal vision-language model that generates spatially grounded and temporally coherent 3D representations to guide policy execution. The smooth spatial masks focus on task-relevant geometry and stabilize latent representations, enabling online replanning and long-horizon reasoning. Experiments on RLBench and real-world manipulation tasks show that \method significantly outperforms state-of-the-art baselines, improving zero-shot success rates by 44.6% and 30.3%. These results demonstrate that offloading spatio-temporal reasoning to VLMs with unified 3D-4D representations substantially improves robustness and generalization for open-world robotic manipulation. Project website: https://oucx117.github.io/ST-VLA/.

cs.RO↗
arXiv2026

AdaClearGrasp: Learning Adaptive Clearing for Zero-Shot Robust Dexterous Grasping in Densely Cluttered Environments

In densely cluttered environments, physical interference, visual occlusions, and unstable contacts often cause direct dexterous grasping to fail, while aggressive singulation strategies may compromise safety. Enabling robots to adaptively decide whether to clear surrounding objects or directly grasp the target is therefore crucial for robust manipulation. We propose AdaClearGrasp, a closed-loop decision-execution framework for adaptive clearing and zero-shot dexterous grasping in densely cluttered environments. The framework formulates manipulation as a controllable high-level decision process that determines whether to directly grasp the target or first clear surrounding objects. A pretrained vision-language model (VLM) interprets visual observations and language task descriptions to reason about grasp interference and generate a high-level planning skeleton, which invokes structured atomic skills through a unified action interface. For dexterous grasping, we train a reinforcement learning policy with a relative hand-object distance representation, enabling zero-shot generalization across diverse object geometries and physical properties. During execution, visual feedback monitors outcomes and triggers replanning upon failures, forming a closed-loop correction mechanism. To evaluate language-conditioned dexterous grasping in clutter, we introduce Clutter-Bench, the first simulation benchmark with graded clutter complexity. It includes seven target objects across three clutter levels, yielding 210 task scenarios. We further perform sim-to-real experiments on three objects under three clutter levels (18 scenarios). Results demonstrate that AdaClearGrasp significantly improves grasp success rates in densely cluttered environments. For more videos and code, please visit our project website: https://chenzixuan99.github.io/adaclear-grasp.github.io/.

cs.RO↗
arXiv2026

MoMaStage: Skill-State Graph Guided Planning and Closed-Loop Execution for Long-Horizon Indoor Mobile Manipulation

Long-horizon indoor mobile manipulation (MoMa) requires robots to execute extended navigation-manipulation sequences whose feasibility depends on state changes induced by preceding skills. Vision-language models (VLMs) can decompose instructions into plausible skill sequences, but they do not reliably track such cumulative embodiment constraints or revise a plan when execution deviates from expectation. We present MoMaStage, a map-light framework for state-consistent planning and closed-loop execution in long-horizon indoor MoMa. MoMaStage couples a frozen VLM with robot execution through three mechanisms: (i) a hierarchical library of grounded, executable skills and a topology-only projection of a Skill-State Graph (SSG) that constrains the VLM's planning space; (ii) an SSG verifier that propagates scene-region and gripper-occupancy state to reject infeasible plans before execution; and (iii) an event-driven monitor that triggers graph-grounded repair only when an observed outcome invalidates the remaining plan. The SSG captures the compact embodiment state needed for skill sequencing without requiring a dense scene map, while geometric and contact-level conditions remain within the underlying controllers. Experiments in physics-rich simulation and on a real mobile manipulator show that MoMaStage improves planning validity and long-horizon execution survival over the evaluated baselines, while reducing latency and model token consumption.

cs.RO↗
arXiv2026

RoTri-Diff: A Spatial Robot-Object Triadic Interaction-Guided Diffusion Model for Bimanual Manipulation

Bimanual manipulation is a fundamental robotic skill that requires continuous and precise coordination between two arms. While imitation learning (IL) is the dominant paradigm for acquiring this capability, existing approaches, whether robot-centric or object-centric, often overlook the dynamic geometric relationship among the two arms and the manipulated object. This limitation frequently leads to inter-arm collisions, unstable grasps, and degraded performance in complex tasks. To address this, in this paper we explicitly models the Robot-Object Triadic Interaction (RoTri) representation in bimanual systems, by encoding the relative 6D poses between the two arms and the object to capture their spatial triadic relationship and establish continuous triangular geometric constraints. Building on this, we further introduce RoTri-Diff, a diffusion-based imitation learning framework that combines RoTri constraints with robot keyposes and object motion in a hierarchical diffusion process. This enables the generation of stable, coordinated trajectories and robust execution across different modes of bimanual manipulation. Extensive experiments show that our approach outperforms state-of-the-art baselines by 10.2% on 11 representative RLBench2 tasks and achieves stable performance on 4 challenging real-world bimanual tasks. Project website: https://rotri-diff.github.io/.

cs.RO↗
arXiv2026

INHerit-SG: Incremental Hierarchical Semantic Scene Graphs with RAG-Style Retrieval

Driven by recent advancements in foundation models, semantic scene graphs have emerged as a promising paradigm for high-level 3D environmental abstraction in robot navigation. However, existing frameworks struggle to successfully handle complex embodied queries while ensuring continuous semantic graph construction. To address these limitations, we present INHerit-SG, an asynchronous dual-stream architecture that systematically structures the 3D environment into a RAG-ready knowledge base. Specifically, our framework integrates comprehensive node representations, an event-triggered asynchronous update scheme, and a structured retrieval mechanism. While geometric segmentation is decoupled from semantic reasoning to maintain mapping efficiency, the semantic nodes also store natural language summaries to support text-based retrieval. Furthermore, we propose an interpretable retrieval pipeline that couples the reasoning capabilities of multi-role LLMs with the topological structure of the scene graph, followed by a visual verification process to mitigate false positives. We evaluate INHerit-SG on a newly constructed benchmark for complex embodied semantic query retrieval, HM3DSem-SQR, and in real-world environments. Experiments demonstrate that our system achieves state-of-the-art performance on complex queries, especially for those involving negations and chained spatial constraints. Project Page: https://fangyuktung.github.io/INHeritSG.github.io/

cs.RO↗
arXiv2026

Thinking-Based Non-Thinking: Solving the Reward Hacking Problem in Training Hybrid Reasoning Models via Reinforcement Learning

Large reasoning models (LRMs) have attracted much attention due to their exceptional performance. However, their performance mainly stems from thinking, a long Chain of Thought (CoT), which significantly increase computational overhead. To address this overthinking problem, existing work focuses on using reinforcement learning (RL) to train hybrid reasoning models that automatically decide whether to engage in thinking or not based on the complexity of the query. Unfortunately, using RL will suffer the the reward hacking problem, e.g., the model engages in thinking but is judged as not doing so, resulting in incorrect rewards. To mitigate this problem, existing works either employ supervised fine-tuning (SFT), which incurs high computational costs, or enforce uniform token limits on non-thinking responses, which yields limited mitigation of the problem. In this paper, we propose Thinking-Based Non-Thinking (TNT). It does not employ SFT, and sets different maximum token usage for responses not using thinking across various queries by leveraging information from the solution component of the responses using thinking. Experiments on five mathematical benchmarks demonstrate that TNT reduces token usage by around 50% compared to DeepSeek-R1-Distill-Qwen-1.5B/7B and DeepScaleR-1.5B, while significantly improving accuracy. In fact, TNT achieves the optimal trade-off between accuracy and efficiency among all tested methods. Additionally, the probability of reward hacking problem in TNT's responses, which are classified as not using thinking, remains below 10% across all tested datasets.

cs.AI↗
arXiv2025

ManiLong-Shot: Interaction-Aware One-Shot Imitation Learning for Long-Horizon Manipulation

One-shot imitation learning (OSIL) offers a promising way to teach robots new skills without large-scale data collection. However, current OSIL methods are primarily limited to short-horizon tasks, thus limiting their applicability to complex, long-horizon manipulations. To address this limitation, we propose ManiLong-Shot, a novel framework that enables effective OSIL for long-horizon prehensile manipulation tasks. ManiLong-Shot structures long-horizon tasks around physical interaction events, reframing the problem as sequencing interaction-aware primitives instead of directly imitating continuous trajectories. This primitive decomposition can be driven by high-level reasoning from a vision-language model (VLM) or by rule-based heuristics derived from robot state changes. For each primitive, ManiLong-Shot predicts invariant regions critical to the interaction, establishes correspondences between the demonstration and the current observation, and computes the target end-effector pose, enabling effective task execution. Extensive simulation experiments show that ManiLong-Shot, trained on only 10 short-horizon tasks, generalizes to 20 unseen long-horizon tasks across three difficulty levels via one-shot imitation, achieving a 22.8% relative improvement over the SOTA. Additionally, real-robot experiments validate ManiLong-Shot's ability to robustly execute three long-horizon manipulation tasks via OSIL, confirming its practical applicability.

cs.RO↗
arXiv2025

Efficient Reinforcement Learning with Semantic and Token Entropy for LLM Reasoning

Reinforcement learning with verifiable rewards (RLVR) has demonstrated superior performance in enhancing the reasoning capability of large language models (LLMs). However, this accuracy-oriented learning paradigm often suffers from entropy collapse, which reduces policy exploration and limits reasoning capabilities. To address this challenge, we propose an efficient reinforcement learning framework that leverages entropy signals at both the semantic and token levels to improve reasoning. From the data perspective, we introduce semantic entropy-guided curriculum learning, organizing training data from low to high semantic entropy to guide progressive optimization from easier to more challenging tasks. For the algorithmic design, we adopt non-uniform token treatment by imposing KL regularization on low-entropy tokens that critically impact policy exploration and applying stronger constraints on high-covariance portions within these tokens. By jointly optimizing data organization and algorithmic design, our method effectively mitigates entropy collapse and enhances LLM reasoning. Experimental results across 6 benchmarks with 3 different parameter-scale base models demonstrate that our method outperforms other entropy-based approaches in improving reasoning.

cs.AI↗
arXiv2025

Multi-Agent Reinforcement Learning with Communication-Constrained Priors

Communication is one of the effective means to improve the learning of cooperative policy in multi-agent systems. However, in most real-world scenarios, lossy communication is a prevalent issue. Existing multi-agent reinforcement learning with communication, due to their limited scalability and robustness, struggles to apply to complex and dynamic real-world environments. To address these challenges, we propose a generalized communication-constrained model to uniformly characterize communication conditions across different scenarios. Based on this, we utilize it as a learning prior to distinguish between lossy and lossless messages for specific scenarios. Additionally, we decouple the impact of lossy and lossless messages on distributed decision-making, drawing on a dual mutual information estimatior, and introduce a communication-constrained multi-agent reinforcement learning framework, quantifying the impact of communication messages into the global reward. Finally, we validate the effectiveness of our approach across several communication-constrained benchmarks.

cs.AI↗
arXiv2025

R2Q: Towards Robust 2-Bit Large Language Models via Residual Refinement Quantization

The rapid progress of Large Language Models (LLMs) has brought substantial computational and memory demands, spurring the adoption of low-bit quantization. While 8-bit and 4-bit formats have become prevalent, extending quantization to 2 bits remains challenging due to severe accuracy degradation. To address this, we propose Residual Refinement Quantization (R2Q)-a novel 2-bit quantization framework that decomposes the process into two sequential 1-bit sub-quantizations, forming an adaptive quantization lattice. Extensive evaluations on Llama, OPT, and Qwen across diverse benchmarks-covering question answering, commonsense reasoning, and language modeling-demonstrate that R2Q consistently outperforms existing 2-bit quantization methods in both fine-grained and coarse-grained settings. By refining quantization through a residual learning mechanism, R2Q enhances performance, improves training stability, and accelerates convergence under extreme compression. Furthermore, its modular design enables seamless integration with existing quantization-aware training (QAT) frameworks.

cs.CL↗