arXiv subjects

Hongyao Tang

Publications and source records attributed to Hongyao Tang.

At least 19 recordsLinked to original sources

arXiv2026

WarpSAC: Towards the Pinnacle of Scalable Off-policy RL by Rethinking Exploration and Exploitation

Massively parallel simulation changes the data regime in which off-policy reinforcement learning (RL) is trained, challenging stabilizers designed for data-limited replay. Through controlled experiments across eight benchmark families, we show that these stabilizers are data-regime-dependent: parameter normalization helps with narrow replay coverage but restricts value fitting when data are abundant, while clipped double-Q can be relaxed in high-throughput manipulation. Age-biased replay weighting improves learning efficiency across regimes, especially with limited network capacity. Based on these findings, we propose WarpSAC, a regime-aware family of off-policy RL algorithms. WarpSAC uses Sample Weight Decay for efficient exploitation and provides two variants: WarpSAC-L (Norm ON, clipped double-Q) for data-limited CPU-scale training, and WarpSAC-A (Norm OFF, single-Q) for data-abundant GPU-parallel training. WarpSAC improves normalized score--step AUC over FlashSAC by 4.5% across nine CPU-scale environments and 23.1% across fourteen GPU-parallel environments. It increases UnitreeG1TransportBox-v1 success rate from 19.8% to 96.4%, improves mean normalized wall-time AUC on MuJoCo Playground by 19.1%, and achieves 36.4% faster sim-to-real deployment on Unitree G1 than FlashSAC. These results show that scalable off-policy RL should adapt its stabilizers to the available data regime.

cs.LG↗

arXiv2026

Towards a Formal Definition of Agent Memory: Basis, Span, Optimality, and the Sequential Memory Problem

Despite the wide deployment of memory in large-model agents, there is no unified formal account of what a memory is or when it is optimal. This paper takes a first step toward this account. The central idea is that memory is a basis, knowledge is its span, and answerability is a coverage problem: an agent stores events extracted from a material; a generation operator turns any event set into the knowledge it entails; and a query is answerable exactly when some single item in the span covers it. The optimal memory is then the capacity-constrained maximizer of expected coverage, and its value traces a utility--capacity frontier, the common yardstick on which memory systems can be compared. Next, we consider noise in the memory and discuss coverage versus precision under it: a memory may store false claims, so the write policy must infer the truth of what it stores. Drawing an analogy with biological memory, which is formed continuously through ongoing experience, we formalize the continual agent-memory problem in a sequential MDP that covers multiple levels, where memory is the state, writing is the action, and the utility settled at query time is the delayed reward that drives learning. To make the framework concrete, we instantiate it on Homer's \emph{Odyssey}, turning the frontier, the compression zone, and the divergence of coverage from precision into concrete numbers. Finally, we position existing systems within the framework, making ``how good is a memory'' measurable and recasting the open problems of constructing and learning agent memory as concrete research questions.

cs.LG↗

arXiv2026

The Mirage of Optimizing Training Policies: Monotonic Inference Policies as the Real Objective for LLM Reinforcement Learning

Reinforcement learning (RL) has gained growing attention in large language model (LLM) post-training, yet RL training remains fragile and can suffer from instability or collapse. One vital cause is training-inference mismatch: LLM adopts separate inference and training engines for generation efficiency and training precision, which in practice exhibits inconsistent probabilities for the same trajectories on training and inference sides, even with synchronized model parameters. This naturally induces a special type of off-policyness ever existing and poisoning the training. Prior works have made various efforts in addressing the off-policyness to stabilize the training policies under the mismatch. In this paper, we point out the objective misalignment neglected by existing works that an effective update to the policy in the training engine not necessarily ensures the improvement of the inference policy, i.e., the one used in deployment. To this end, we propose a new policy optimization objective for LLM RL, named Monotonic Inference Policy Improvement (MIPI). Following this principle, we introduce Monotonic Inference Policy Update (MIPU), a two-step LLM RL framework that constructs sampler-referenced candidate updates and selectively accepts synchronized candidates using an inference-side gap proxy. Experiments conducted on two model scales under high mismatch show that MIPU improves average reasoning performance and training stability.

cs.LG↗

arXiv2026

RoboPIN: Grounded Embodied Reasoning via Pinned Chain-of-Thought

Embodied reasoning requires models to perceive task-relevant objects and spaces in physical environments and maintain consistent visual grounding throughout multi-step reasoning. However, current vision-language models rely on text-only or coordinate-augmented chain-of-thought, where entity references remain implicit and ambiguous. This may cause the reasoning process to decouple from visual evidence, entity references to drift across steps, and a causal disconnection between the reasoning trajectory and the final answer, with these problems further amplified in multi-view scenarios due to cross-view appearance changes. To address these issues, we propose Pinned Chain-of-Thought (PinCoT), a structured reasoning paradigm that pins every reasoning step to visual evidence. PinCoT introduces the concept of reasoning anchor, which binds each task-relevant entity to a structured visual anchor with entity name, unique identity, view index, and spatial grounding, enabling consistent entity tracking across reasoning steps and views. We build a fully automated data generation pipeline to construct PIN-170K, a high-quality PinCoT-formatted reasoning dataset. We then train RoboPIN through three-stage post-training that progressively injects embodied knowledge, structured reasoning ability, and process-supervised alignment, with rewards that directly constrain both anchor localization and identity consistency during reasoning. On 14 benchmarks covering embodied spatial reasoning, multi-view reasoning, and pointing, RoboPIN with only 4B parameters surpasses 7B level open-source embodied models on average, achieving a 12% average improvement over the strongest 7B baseline, Mimo-Embodied. Further analysis shows that PinCoT improves grounding accuracy and cross-step identity consistency, validating the effectiveness of process supervision.

cs.AI↗

arXiv2026

Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models

We introduce Embodied-R1.5, a unified Embodied Foundation Model (EFM) that integrates comprehensive embodied reasoning capabilities, spanning embodied cognition, task planning, correction, and pointing, within a single architecture toward general physical intelligence. Leveraging three automated data construction pipelines to significantly expand the data coverage of critical capabilities, we build a large-scale data system of over 15B tokens, and design a multi-task balanced RL recipe to alleviate heterogeneous task conflicts. We further introduce a Planner-Grounder-Corrector (PGC) closed-loop framework that enables a single model to autonomously execute and self-correct over long-horizon tasks. With only 8B parameters, Embodied-R1.5 achieves SOTA on 16 out of 24 embodied VLM benchmarks, surpassing leading models like Gemini-Robotics-ER-1.5 and GPT-5.4. Benefiting from the internalized embodied capabilities, Embodied-R1.5 can be fine-tuned into a VLA with only a small amount of data, outperforming leading VLA models like $\pi_{0.5}$ across 4 popular manipulation benchmark suites. We further conduct extensive zero-shot real-robot experiments, validating performance in instruction following, affordance grounding, articulated object manipulation, and long-horizon complex tasks, demonstrating strong generalization to the physical world. We open-source model weights, datasets, training code, and EmbodiedEvalKit, an evaluation framework tailored for embodied tasks, to facilitate future research in EFMs.

cs.RO↗

arXiv2026

Reformulate LLM Reinforcement Learning for Efficient Training under Black-box Discrepancy

Reinforcement Learning (RL) has emerged as a pivotal post-training paradigm, yet it frequently suffers from unpredictable sub-optimum performance or even training collapses. Recent findings attribute these failures to a hidden train-inference discrepancy (or mismatch), stemming from the disparate underlying engines and architecture. We find that the training policy can actively self-correct such a discrepancy when provided with an appropriate learning signal. Then, we further empirically identify a discrepancy tolerance region: within this region, aggressively narrowing the discrepancy can suppress policy exploration and reduce learning efficiency, whereas outside this region, reducing excessive discrepancy improves optimization consistency and raises the achievable local performance ceiling. According to such findings, we formulate this problem as a Discrepancy-Constrained Markov Decision Process (DCMDP), where reward maximization is coupled with a constraint that aligns training-Inference behavior, achieving stable dual-objective optimization. To adaptively balance performance improvement and discrepancy control, we introduce a Lagrangian relaxation mechanism that dynamically adjusts the relative weight of the two objectives according to the current degree of discrepancy violation. This enables stable dual-objective optimization: the policy is allowed to explore freely within the tolerance region, while being guided back when the discrepancy exceeds the safe boundary. Empirically, DCMDP significantly improves the performance of 8B dense model (Qwen-3-8b) and 30B Mixture-of-Expert model (Qwen-3-30bA3b), and enables a heterogeneous training paradigm, where LLMs can be optimized in high-fidelity training setup while being explicitly aligned for low-cost, resource-constrained inference deployment.

cs.LG↗

arXiv2026

ForceFlow: Learning to Feel and Act via Contact-Driven Flow Matching

Existing imitation learning methods enable robots to interact autonomously with the physical environment. However, contact-rich manipulation tasks remain a significant challenge due to complex contact dynamics that demand high-precision force feedback and control. Although recent efforts have attempted to integrate force/torque sensing into policies, how to build a simple yet effective framework that achieves robust generalization under multimodal observations remains an open question. In this paper, we propose ForceFlow, a force-aware reactive framework built upon flow matching. For contact-stage policy design, we investigate force signal fusion mechanisms and adopt an asymmetric multimodal fusion architecture that treats force as a global regulatory signal, combined with a joint prediction paradigm that enhances the policy's understanding of instantaneous force and historical information, thereby achieving deep coupling between force and motion. For task-level hierarchical decomposition, we divide manipulation into a vision-dominant approach stage (VLM-based pointing for target localization) and a touch-dominant interaction stage (force-driven contact execution), with a Vision-to-Force (V2F) handover mechanism that explicitly decouples spatial generalization from contact regulation. Experimental results across six real-world contact-rich tasks demonstrate that ForceFlow achieves a 37% success rate improvement over the strong baseline ForceVLA while maintaining significantly lower cost. Moreover, ForceFlow exhibits accurate force signal prediction and demonstrates superior performance in contact force self-regulation and zero-shot out-of-distribution (OOD) generalization.

cs.RO↗

arXiv2026

The Rank and Gradient Lost in Non-stationarity: Sample Weight Decay for Mitigating Plasticity Loss in Reinforcement Learning

Deep reinforcement learning (RL) suffers from plasticity loss severely due to the nature of non-stationarity, which impairs the ability to adapt to new data and learn continually. Unfortunately, our understanding of how plasticity loss arises, dissipates, and can be dissolved remains limited to empirical findings, leaving the theoretical end underexplored.To address this gap, we study the plasticity loss problem from the theoretical perspective of network optimization. By formally characterizing the two culprit factors in online RL process: the non-stationarity of data distributions and the non-stationarity of targets induced by bootstrapping, our theory attributes the loss of plasticity to two mechanisms: the rank collapse of the Neural Tangent Kernel (NTK) Gram matrix and the $\Theta(\frac{1}{k})$ decay of gradient magnitude. The first mechanism echoes prior empirical findings from the theoretical perspective and sheds light on the effects of existing methods, e.g., network reset, neuron recycle, and noise injection. Against this backdrop, we focus primarily on the second mechanism and aim to alleviate plasticity loss by addressing the gradient attenuation issue, which is orthogonal to existing methods. We propose Sample Weight Decay -- a lightweight method to restore gradient magnitude, as a general remedy to plasticity loss for deep RL methods based on experience replay. In experiments, we evaluate the efficacy of \methodName upon TD3, \myadded{Double DQN} and SAC with SimBa architecture in MuJoCo, \myadded{ALE} and DeepMind Control Suite tasks. The results demonstrate that \methodName effectively alleviates plasticity loss and consistently improves learning performance across various configurations of deep RL algorithms, UTD, network architectures, and environments, achieving SOTA performance on challenging DMC Humanoid tasks.

cs.LG↗

arXiv2025

MUVLA: Learning to Explore Object Navigation via Map Understanding

In this paper, we present MUVLA, a Map Understanding Vision-Language-Action model tailored for object navigation. It leverages semantic map abstractions to unify and structure historical information, encoding spatial context in a compact and consistent form. MUVLA takes the current and history observations, as well as the semantic map, as inputs and predicts the action sequence based on the description of goal object. Furthermore, it amplifies supervision through reward-guided return modeling based on dense short-horizon progress signals, enabling the model to develop a detailed understanding of action value for reward maximization. MUVLA employs a three-stage training pipeline: learning map-level spatial understanding, imitating behaviors from mixed-quality demonstrations, and reward amplification. This strategy allows MUVLA to unify diverse demonstrations into a robust spatial representation and generate more rational exploration strategies. Experiments on HM3D and Gibson benchmarks demonstrate that MUVLA achieves great generalization and learns effective exploration behaviors even from low-quality or partially successful trajectories.

cs.RO↗

arXiv2025

Embodied Arena: A Comprehensive, Unified, and Evolving Evaluation Platform for Embodied AI

Embodied AI development significantly lags behind large foundation models due to three critical challenges: (1) lack of systematic understanding of core capabilities needed for Embodied AI, making research lack clear objectives; (2) absence of unified and standardized evaluation systems, rendering cross-benchmark evaluation infeasible; and (3) underdeveloped automated and scalable acquisition methods for embodied data, creating critical bottlenecks for model scaling. To address these obstacles, we present Embodied Arena, a comprehensive, unified, and evolving evaluation platform for Embodied AI. Our platform establishes a systematic embodied capability taxonomy spanning three levels (perception, reasoning, task execution), seven core capabilities, and 25 fine-grained dimensions, enabling unified evaluation with systematic research objectives. We introduce a standardized evaluation system built upon unified infrastructure supporting flexible integration of 22 diverse benchmarks across three domains (2D/3D Embodied Q&A, Navigation, Task Planning) and 30+ advanced models from 20+ worldwide institutes. Additionally, we develop a novel LLM-driven automated generation pipeline ensuring scalable embodied evaluation data with continuous evolution for diversity and comprehensiveness. Embodied Arena publishes three real-time leaderboards (Embodied Q&A, Navigation, Task Planning) with dual perspectives (benchmark view and capability view), providing comprehensive overviews of advanced model capabilities. Especially, we present nine findings summarized from the evaluation results on the leaderboards of Embodied Arena. This helps to establish clear research veins and pinpoint critical research problems, thereby driving forward progress in the field of Embodied AI.

cs.RO↗

arXiv2025

Embodied-R1: Reinforced Embodied Reasoning for General Robotic Manipulation

Generalization in embodied AI is hindered by the "seeing-to-doing gap," which stems from data scarcity and embodiment heterogeneity. To address this, we pioneer "pointing" as a unified, embodiment-agnostic intermediate representation, defining four core embodied pointing abilities that bridge high-level vision-language comprehension with low-level action primitives. We introduce Embodied-R1, a 3B Vision-Language Model (VLM) specifically designed for embodied reasoning and pointing. We use a wide range of embodied and general visual reasoning datasets as sources to construct a large-scale dataset, Embodied-Points-200K, which supports key embodied pointing capabilities. We then train Embodied-R1 using a two-stage Reinforced Fine-tuning (RFT) curriculum with a specialized multi-task reward design. Embodied-R1 achieves state-of-the-art performance on 11 embodied spatial and pointing benchmarks. Critically, it demonstrates robust zero-shot generalization by achieving a 56.2% success rate in the SIMPLEREnv and 87.5% across 8 real-world XArm tasks without any task-specific fine-tuning, representing a 62% improvement over strong baselines. Furthermore, the model exhibits high robustness against diverse visual disturbances. Our work shows that a pointing-centric representation, combined with an RFT training paradigm, offers an effective and generalizable pathway to closing the perception-action gap in robotics.

cs.RO↗

arXiv2025

Scaling DRL for Decision Making: A Survey on Data, Network, and Training Budget Strategies

In recent years, the expansion of neural network models and training data has driven remarkable progress in deep learning, particularly in computer vision and natural language processing. This advancement is underpinned by the concept of Scaling Laws, which demonstrates that scaling model parameters and training data enhances learning performance. While these fields have witnessed breakthroughs, such as the development of large language models like GPT-4 and advanced vision models like Midjourney, the application of scaling laws in deep reinforcement learning (DRL) remains relatively unexplored. Despite its potential to improve performance, the integration of scaling laws into DRL for decision making has not been fully realized. This review addresses this gap by systematically analyzing scaling strategies in three dimensions: data, network, and training budget. In data scaling, we explore methods to optimize data efficiency through parallel sampling and data generation, examining the relationship between data volume and learning outcomes. For network scaling, we investigate architectural enhancements, including monolithic expansions, ensemble and MoE methods, and agent number scaling techniques, which collectively enhance model expressivity while posing unique computational challenges. Lastly, in training budget scaling, we evaluate the impact of distributed training, high replay ratios, large batch sizes, and auxiliary training on training efficiency and convergence. By synthesizing these strategies, this review not only highlights their synergistic roles in advancing DRL for decision making but also provides a roadmap for future research. We emphasize the importance of balancing scalability with computational efficiency and outline promising directions for leveraging scaling to unlock the full potential of DRL in various tasks such as robot control, autonomous driving and LLM training.

cs.LG↗

arXiv2025

Efficient Morphology-Aware Policy Transfer to New Embodiments

Morphology-aware policy learning is a means of enhancing policy sample efficiency by aggregating data from multiple agents. These types of policies have previously been shown to help generalize over dynamic, kinematic, and limb configuration variations between agent morphologies. Unfortunately, these policies still have sub-optimal zero-shot performance compared to end-to-end finetuning on morphologies at deployment. This limitation has ramifications in practical applications such as robotics because further data collection to perform end-to-end finetuning can be computationally expensive. In this work, we investigate combining morphology-aware pretraining with parameter efficient finetuning (PEFT) techniques to help reduce the learnable parameters necessary to specialize a morphology-aware policy to a target embodiment. We compare directly tuning sub-sets of model weights, input learnable adapters, and prefix tuning techniques for online finetuning. Our analysis reveals that PEFT techniques in conjunction with policy pre-training generally help reduce the number of samples to necessary to improve a policy compared to training models end-to-end from scratch. We further find that tuning as few as less than 1% of total parameters will improve policy performance compared the zero-shot performance of the base pretrained a policy.

cs.LG↗

arXiv2025

Squeeze the Soaked Sponge: Efficient Off-policy Reinforcement Finetuning for Large Language Model

Reinforcement Learning (RL) has demonstrated its potential to improve the reasoning ability of Large Language Models (LLMs). One major limitation of most existing Reinforcement Finetuning (RFT) methods is that they are on-policy RL in nature, i.e., data generated during the past learning process is not fully utilized. This inevitably comes at a significant cost of compute and time, posing a stringent bottleneck on continuing economic and efficient scaling. To this end, we launch the renaissance of off-policy RL and propose Reincarnating Mix-policy Proximal Policy Gradient (ReMix), a general approach to enable on-policy RFT methods like PPO and GRPO to leverage off-policy data. ReMix consists of three major components: (1) Mix-policy proximal policy gradient with an increased Update-To-Data (UTD) ratio for efficient training; (2) KL-Convex policy constraint to balance the trade-off between stability and flexibility; (3) Policy reincarnation to achieve a seamless transition from efficient early-stage learning to steady asymptotic improvement. In our experiments, we train a series of ReMix models upon PPO, GRPO and 1.5B, 7B base models. ReMix shows an average Pass@1 accuracy of 52.10% (for 1.5B model) with 0.079M response rollouts, 350 training steps and achieves 63.27%/64.39% (for 7B model) with 0.007M/0.011M response rollouts, 50/75 training steps, on five math reasoning benchmarks (i.e., AIME'24, AMC'23, Minerva, OlympiadBench, and MATH500). Compared with 15 recent advanced models, ReMix shows SOTA-level performance with an over 30x to 450x reduction in training cost in terms of rollout data volume. In addition, we reveal insightful findings via multifaceted analysis, including the implicit preference for shorter responses due to the Whipping Effect of off-policy discrepancy, the collapse mode of self-reflection behavior under the presence of severe off-policyness, etc.

cs.LG↗

arXiv2025

Mitigating Plasticity Loss in Continual Reinforcement Learning by Reducing Churn

Plasticity, or the ability of an agent to adapt to new tasks, environments, or distributions, is crucial for continual learning. In this paper, we study the loss of plasticity in deep continual RL from the lens of churn: network output variability for out-of-batch data induced by mini-batch training. We demonstrate that (1) the loss of plasticity is accompanied by the exacerbation of churn due to the gradual rank decrease of the Neural Tangent Kernel (NTK) matrix; (2) reducing churn helps prevent rank collapse and adjusts the step size of regular RL gradients adaptively. Moreover, we introduce Continual Churn Approximated Reduction (C-CHAIN) and demonstrate it improves learning performance and outperforms baselines in a diverse range of continual learning environments on OpenAI Gym Control, ProcGen, DeepMind Control Suite, and MinAtar benchmarks.

cs.LG↗

arXiv2025

Can We Optimize Deep RL Policy Weights as Trajectory Modeling?

Learning the optimal policy from a random network initialization is the theme of deep Reinforcement Learning (RL). As the scale of DRL training increases, treating DRL policy network weights as a new data modality and exploring the potential becomes appealing and possible. In this work, we focus on the policy learning path in deep RL, represented by the trajectory of network weights of historical policies, which reflects the evolvement of the policy learning process. Taking the idea of trajectory modeling with Transformer, we propose Transformer as Implicit Policy Learner (TIPL), which processes policy network weights in an autoregressive manner. We collect the policy learning path data by running independent RL training trials, with which we then train our TIPL model. In the experiments, we demonstrate that TIPL is able to fit the implicit dynamics of policy learning and perform the optimization of policy network by inference.

cs.LG↗

arXiv2025

Dual Ensembled Multiagent Q-Learning with Hypernet Regularizer

Overestimation in single-agent reinforcement learning has been extensively studied. In contrast, overestimation in the multiagent setting has received comparatively little attention although it increases with the number of agents and leads to severe learning instability. Previous works concentrate on reducing overestimation in the estimation process of target Q-value. They ignore the follow-up optimization process of online Q-network, thus making it hard to fully address the complex multiagent overestimation problem. To solve this challenge, in this study, we first establish an iterative estimation-optimization analysis framework for multiagent value-mixing Q-learning. Our analysis reveals that multiagent overestimation not only comes from the computation of target Q-value but also accumulates in the online Q-network's optimization. Motivated by it, we propose the Dual Ensembled Multiagent Q-Learning with Hypernet Regularizer algorithm to tackle multiagent overestimation from two aspects. First, we extend the random ensemble technique into the estimation of target individual and global Q-values to derive a lower update target. Second, we propose a novel hypernet regularizer on hypernetwork weights and biases to constrain the optimization of online global Q-network to prevent overestimation accumulation. Extensive experiments in MPE and SMAC show that the proposed method successfully addresses overestimation across various tasks.

cs.MA↗

arXiv2024

Improving Deep Reinforcement Learning by Reducing the Chain Effect of Value and Policy Churn

Deep neural networks provide Reinforcement Learning (RL) powerful function approximators to address large-scale decision-making problems. However, these approximators introduce challenges due to the non-stationary nature of RL training. One source of the challenges in RL is that output predictions can churn, leading to uncontrolled changes after each batch update for states not included in the batch. Although such a churn phenomenon exists in each step of network training, how churn occurs and impacts RL remains under-explored. In this work, we start by characterizing churn in a view of Generalized Policy Iteration with function approximation, and we discover a chain effect of churn that leads to a cycle where the churns in value estimation and policy improvement compound and bias the learning dynamics throughout the iteration. Further, we concretize the study and focus on the learning issues caused by the chain effect in different settings, including greedy action deviation in value-based methods, trust region violation in proximal policy optimization, and dual bias of policy value in actor-critic methods. We then propose a method to reduce the chain effect across different settings, called Churn Approximated ReductIoN (CHAIN), which can be easily plugged into most existing DRL algorithms. Our experiments demonstrate the effectiveness of our method in both reducing churn and improving learning performance across online and offline, value-based and policy-based RL settings, as well as a scaling setting.

cs.LG↗