arXiv subjects

Yang Liu

Publications and source records attributed to Yang Liu.

At least 91 records · Page 5Linked to original sources

arXiv2026

PEARL: Process-Experience Aware Reinforcement Learning for Multi-Turn Agents

Terminal rewards make multi-turn agents trainable without step-level annotations, but group-relative advantages provide little guidance about which intermediate decisions to adjust. Failed trajectories also contain reasonable actions, while unfavorable decision patterns can recur as training proceeds. We introduce PEARL (Process-Experience Aware Reinforcement Learning), which improves local credit assignment without an additional critic. Given two continuations from the same history with different outcomes, PEARL locates their first divergence and builds a feature representation of each branch's local decision. It clusters these decision features in recent failure and success memories to identify recurring local patterns. Using this pattern memory, PEARL selectively strengthens negative advantages at lower-reward decisions, while leaving positive advantages and the base policy optimizer unchanged. On AppWorld, PEARL improves over ARPO on Normal and Challenge tasks at both 4B and 8B scales. On ALFWorld, it achieves 97.9% mean success at 4B, 2.7 points above the strongest baseline, while remaining competitive at 8B. Ablations of shaping location, success-bank filtering, and memory window support the value of selective, recurring-pattern-based updates.

cs.AI↗

arXiv2026

Predict in Chunks, Revise in Spans: History-conditioned Action Regeneration for Robotic Manipulation

Action chunking improves temporal coherence in visuomotor imitation learning by jointly predicting sequences of future actions. However, many existing policies regenerate entire action chunks even when errors are localized, potentially replacing otherwise valid predictions. We introduce Contextual Masked Action Policy (CMAP), a visuomotor policy that refines continuous action chunks through history-conditioned local regeneration. Specifically, CMAP introduces a history-conditioned action tape that combines executed commands with masked future positions to produce coarse action predictions and per-step uncertainty scores. An uncertainty-guided span regeneration mechanism then selects a high-uncertainty contiguous span and reconstructs it using the executed history and retained future predictions as context. By jointly reconstructing neighboring actions, CMAP models local temporal dependencies while preserving all predictions outside the selected span. Extensive experiments on nine simulated and four real-world manipulation tasks demonstrate that CMAP achieves superior overall performance over several competitive baselines. Ablation studies validate the complementary contributions of its key components, while held-out trajectory analysis shows that local regeneration reduces prediction errors within selected spans.

cs.RO↗

arXiv2026

PyCache Trap: The Inspection-Execution Gap in Agent Skill Scanners

Agent skills combine instructions with executable resources, giving third-party packages access to an agent's runtime. Existing skill scanners inspect documentation and visible source, but Python may execute a bundled bytecode cache with different behavior. We study this gap between inspection and execution through PyCache Trap, which pairs benign source with a substituted cache accepted by the loader and connects it to a task-relevant invocation. Scanner-guided rewriting changes the invocation wording while preserving the cache body, separating package admission from recognition of the concealed behavior. Across 100 skills and seven scanners, PyCache Trap achieves 94-100% attack success, with no semantic recognition of the cache-resident behavior. We propose execution-aware validation (EAV) to connect inspected instructions, scripts, imports, and runtime artifacts in a typed execution graph. EAV combines grounded behavioral analysis with trusted reproduction of compiled artifacts. It detects all 100 evaluated source-present cache substitutions and reaches 92.8% Recall at 10.0% FPR across five attack families and 200 benign skills. The results support checking the executable artifacts a runtime can select as part of skill admission, within the supported loaders and code-object normalization. The code is released at https://github.com/leo0481/PyCacheTrap.

cs.CR↗

arXiv2026

Diffusion Precipitation Generation with Medium-range Weather Forecasting

Accurate daily precipitation forecasts are important for weather-sensitive decisions. Although recent Weather Foundation Models (WFMs) have achieved significant progress, they are trained on reanalysis data, which exhibit regional and intensity-dependent biases. Since retraining WFMs on observation-based targets is costly, we study Dynamics-Precipitation Decoupling, a plug-and-play framework that uses a separately trained generator to map non-precipitation forecasts from a frozen weather model to observation-aligned daily precipitation. Its key challenges are to capture heterogeneous rainfall distributions and align daily accumulations with atmospheric signals that evolve across space and time. To address them, we train a diffusion generator conditional on atmospheric states and satellite-gauge estimates of precipitation. A multi-scale recurrent encoder summarizes atmospheric evolution, while cross-attention aligns atmospheric and precipitation features during denoising. Extensive evaluation of various backbones, including Pangu-Weather, GraphCast, and HRES, on global and regional forecasting shows improved daily precipitation accuracy and event detection, with better representation of localized accumulation patterns. Global ensemble evaluation shows lower Brier scores than ENS and training-from-scratch GenCast at all reported lead times, with lower CRPS in most comparisons. Remarkably, a heavy-rainfall case study using 2,288 stations in China further supports improved daily accumulation patterns and event detection against ground observations.

cs.LG↗

arXiv2026

PILOT: Guiding Long-Horizon World-Model Planning for Robot Manipulation with Visual Subgoals and Policy Priors

Latent world-model planning has shown promising results in goal-conditioned robotic manipulation, but remains challenging for long-horizon planning due to limited temporal guidance and inefficient action search. A distant visual goal provides little guidance for preparatory actions, while uninformed action sampling makes local optimization inefficient. We propose PILOT (Policy-Initialized Local Optimization with Temporal subgoals), which uses a video model to generate intermediate visual subgoals and a learned policy to initialize the cross-entropy method (CEM). CEM refines candidate action sequences toward each subgoal using an action-conditioned world model in a receding-horizon control loop. In our primary real-robot evaluation across four UR5e manipulation tasks, PILOT achieves 68.8% success, compared with 55.0% without policy initialization, 13.8% for policy-only execution, and 0.0% for direct final-goal planning. Simulation results show that these gains persist over longer horizons: PILOT exceeds direct execution of the same policy by 30 percentage points on long Push-T tasks and by 22 percentage points on LIBERO-10. Policy initialization also accelerates local optimization: on Push-T, CEM matches the cost reached by generic initialization after 10 iterations in only 4 iterations and reduces the final planning cost by 24.1% under the same sampling budget. These results show that combining temporal subgoals with policy-guided action search improves both long-horizon task execution and local planning efficiency. Videos are available on the webpage: https://pilot-planner.github.io/.

cs.RO↗

arXiv2026

DISCERN: Identifiable Disease Evidence for Generalizable Medical Multimodal Learning

Clinical multimodal models can learn enrollment shortcuts shared across modalities and fail on new cohorts. We model these data through a latent structural causal model: disease state and background factors generate physiology through shared mechanisms, while cohorts differ in selection and measurement. Holding acquisition fixed, we identify conditional disease evidence without observing disease-state labels or recovering the full causal model. In common Gaussian coordinates, cross-cohort state alignment and natural-parameter changes identify a minimal selection statistic; conditioning on it removes enrollment-dependent evidence and yields an identifiable reference posterior within the specified shift family. State-dependent interactions can carry evidence invisible to individual-modality posteriors, motivating joint, information-preserving representations. DISCERN combines an information-lattice encoder, a latent-state model, and an outcome head through joint training. Under reversed synthetic enrollment on known latent coordinates, conditioning raises state AUROC from 0.043 to 0.883 and from 0.265 to 0.825, while random conditioning does not. Exploratory cross-database mortality experiments between eICU and MIMIC-IV improve AUROC over matched empirical risk minimization by 0.036 and 0.035. Learned-coordinate diagnostics distinguish preservation of state information from identification of the coordinates required by the estimator.

cs.LG↗

arXiv2026

\$OneMillion-Bench: How Far are Language Agents from Human Experts?

As language models (LMs) evolve from chat assistants to long-horizon agents capable of multi-step reasoning and tool use, existing benchmarks remain largely confined to structured or exam-style tasks that fall short of real-world professional demands. To this end, we introduce \$OneMillion-Bench (\$OMB), a benchmark of 400 expert-curated tasks spanning Law, Finance, Industry, Healthcare, and Natural Science, built to evaluate agents across economically consequential scenarios. Unlike prior work, the benchmark requires retrieving authoritative sources, resolving conflicting evidence, applying domain-specific rules, and making constraint decisions, where correctness depends as much on the reasoning process as the final answer. We adopt a rubric-based evaluation protocol scoring factual accuracy, logical coherence, practical feasibility, and professional compliance, focusing on expert-level problems to ensure meaningful differentiation across agents. Together, \$OMB provides a unified testbed for assessing agentic reliability, professional depth, and an indicator of practical readiness in domain-intensive scenarios.

cs.LG↗

arXiv2026

Learning Kilometer-Scale Weather Prediction with Global-Regional Alignment

Kilometer-scale regional weather forecasting is essential for local weather warnings and weather-sensitive decisions. Existing data-driven approaches often rely on numerical forecasts for large-scale guidance or require additional training of global forecasting components. Pretrained global weather models offer an efficient source of large-scale forecasts, motivating their reuse to guide high-resolution regional prediction. However, this coupling requires aligning global and regional representations across different grids and integrating global guidance with local interactions to advance regional states. We propose ScaleCast, a regional forecasting framework that addresses these challenges through Global-Regional Alignment. Its Global-Regional Conversion module aligns joint global and regional representations with regional locations, while the Global-Regional Alignment and Dynamics block combines aligned guidance with regional neighborhood interactions. Experiments using ERA5 global analyses on a 0.25-degree grid and CERRA regional reanalysis at 5.5 km spacing demonstrate improved regional forecasts across surface and upper-air variables, with a single trained model supporting multiple global forecast drivers (i.e., Pangu-Weather, GraphCast, and HRES) without specific retraining. Fine-tuning on HRRR at 3 km spacing further demonstrates the framework's adaptability to a different regional domain and spatial resolution. Windstorm case studies show improved cyclone positioning and core-pressure estimates, while comparisons with HadISD station observations show closer agreement with local temperature and humidity changes.

cs.LG↗

arXiv2026

Median randomized quasi-Monte Carlo for nested integration

Nested integrals $\int f\left(\int g(\boldsymbol{y},\boldsymbol{x})\mathrm{d}{}\boldsymbol{x}\right)\mathrm{d}{}\boldsymbol{y}$ where $\boldsymbol{y}\in[0,1]^{d_1}$ and $\boldsymbol{x}\in[0,1]^{d_2}$ frequently arise in Bayesian experimental design, medical decision making, and computational finance. Statistical estimators for nested integrals are more computationally demanding than for standard integrals. Recent studies demonstrate that the median randomized quasi-Monte Carlo (mRQMC) method achieves super-polynomial convergence rates for analytic integrands. We extend mRQMC methods to nested integrations and propose a novel double-loop mRQMC estimator with super-polynomial convergence rate if both the inner and outer integrands are analytic. Established estimators for nested integrals, such as double-loop or multilevel Monte Carlo, are by contrast limited to algebraic convergence rates. However, expected information gain (EIG) applications from experimental design typically do not fulfill this condition due to a lack of regularity in the outer integrand. We address this through a theoretically-grounded importance sampling regularization scheme. Although super-polynomial convergence rates remain achievable for the inner integral, only the algebraic rate $α+3/2-ε$ for any $ε>0$ can be theoretically guaranteed for the outer integral in the EIG, where $α\in\mathbb{N}\cup\{0\}$ denotes the order of the mixed derivatives of the regularized outer integrand. Our approach shows substantial improvements over double-loop randomized quasi-Monte Carlo methods based on the sample mean in moderately high-dimensional settings for a simple polynomial example. Moreover, we demonstrate improved convergence behavior of the median-based estimator for the more challenging setting of EIG estimation.

math.NA↗

arXiv2026

Sparse-View 4D Gaussian Splatting via Spatiotemporal Priors and Generative Assistance

We present a 4D Gaussian Splatting framework for the Sparse-View Track of the SIGGRAPH Asia 2026 Volumetric Video Challenge, which requires dynamic scene reconstruction from only six cameras with wide baselines. To achieve robust dynamic reconstruction under such sparse views, our framework integrates three components. (1) Region-adaptive spatial priors: We use foreground masks to guide Gaussian initialization and mask voting to control densification separately for the dynamic foreground and static background. Background geometry is regularized using monocular depth aligned to metric scale. (2) Motion-consistent temporal priors: We provide supervision at intermediate times through frame interpolation and constrain projected Gaussian motion with estimated optical flow. (3) Generative assistance: We place virtual cameras in the widest angular gaps and restore their rendered images using a diffusion-based model conditioned on camera poses. The restored images are iteratively incorporated into training as pseudo-supervision. On the validation set, our framework improves full-frame PSNR from 25.60 dB for the baseline to 29.75 dB. On the official test benchmark, it achieves 30.04 dB full-frame PSNR and 27.88 dB foreground PSNR, ranking first overall in the Sparse-View Track.

cs.CV↗

arXiv2026

RoboRender: Robot-Oriented Video Generation for Visual Sim-to-Real Transfer

Simulation enables large-scale, low-cost robot data generation, but policies trained in simulation often fail to transfer to the real world due to the sim-to-real visual discrepancies. Existing approaches often rely on intermediate representations, which can discard rich semantic information or require additional perception modules at deployment. We address this visual sim-to-real gap with RoboRender, a framework that converts simulated trajectories into photorealistic RGB videos for policy learning. RoboRender trains a robot-oriented video generation model conditioned on simulated depth videos, language instructions, and robot RGB mask videos, preserving simulator geometry, robot motion, and action labels while synthesizing realistic textures, backgrounds, and distractors. The generated RGB videos are paired with simulator-provided states and actions to train policies for zero-shot real-world deployment. On robot video test sets, our video model outperforms depth-conditioned video generation baselines in generation quality. In real-world experiments across pick-and-place, articulated-object manipulation, and mobile manipulation tasks, policies trained on RoboRender-generated data achieve a 71% average success rate, outperforming raw simulation renderings and conventional visual domain randomization by approximately 7.1x and 3.6x, respectively. We further show that policy performance improves with more generated videos per simulation trajectory, increasing opening-task success by 65 percentage points. These results demonstrate that generative video rendering mitigates the visual sim-to-real gap for zero-shot policy transfer. Project website: https://robo-render.github.io/.

cs.RO↗

arXiv2026

Robust distortion riskmetrics under Wasserstein ambiguity

Risk evaluation under distributional ambiguity is central to decision making in finance, economics, and operations research. Wasserstein balls provide a natural way to describe uncertainty around a reference distribution. We solve a natural yet open problem of robust optimization for the class of distortion riskmetrics with Wasserstein distance as the sole ambiguity constraint. This chosen objective class does not require convexity, monotonicity, and continuity of distortion functions, encompassing many common risk measures and deviation measures. First, we characterize conditions under which direct convexification preserves the worst-case value. Second, we develop a constructive method for exact worst-case evaluation when the direct convexification conditions fail. Third, we construct explicit approximate worst-case distributions and provide computable error bounds to assess their accuracy without solving the exact problem. We apply these results to distributionally robust portfolio selection and use numerical experiments to assess approximation accuracy and the resulting portfolio decisions.

q-fin.MF↗

arXiv2026

Robotic Boomerang Throwing via Model-Based Release Design

Throwing objects that generate aerodynamic lift can greatly extend robot throwing beyond ballistic flight. A returning boomerang is a challenging example because its flight depends strongly on the release velocity, attitude, and spin, while robotic manipulators cannot readily reproduce the rapid motions used in human throwing. We present a model-based framework for robotic boomerang throwing centered on the release state. We identify the boomerang flight dynamics in stages to predict how flight changes across design variations. To systematically design the robot throwing motion, we screen candidate parameters according to how strongly and consistently they control release spin under uncertain contact conditions. These models are then used to design the throwing motion and boomerang for a 6-DoF manipulator with limited joint speeds. To our knowledge, this is the first robotic manipulator to generate a returning boomerang flight. In the demonstrated returning trial, the boomerang is released at 51 rad/s (8.1 rev/s), reaches 2.03 m from the robot base, and returns to touch down 0.31 m from the base. The successful release differs significantly from the measured human throws, showing that a robot need not imitate human throwing motion to achieve a returning flight. The project page is available at https://robot-boomerang.github.io

cs.RO↗

arXiv2026

Fast Planning for Multi-object Multi-target Throwing

Robot throwing has emerged as a promising technique for improving efficiency in logistics and warehouse automation, by enlarging the workspace and speeding up the process. To significantly increase the throwing system's throughput, we develop strategies for throwing multiple objects in one swipe. Such multi-object multi-target throwing (MOMT) leverages the large degrees of freedom of anthropomorphic hands. The key is to quickly generate fast and feasible throwing motions, which involves a complex trade-off between short trajectory duration and short planning time. We solve this problem in two stages. Offline, we build a model of the feasible set by combining object's inverted flying dynamics and the robot's kinematics and dynamics. Online, we generate feasible throws through fast solution matching and filtering of object's valid detach state and robot's feasible state that can compose sequences of throws in less than 5 ms. We validate the framework on a 7-DoF manipulator equipped with a multi-fingered hand. In simulation, coordinated two-object throwing reduces execution time by up to 46% compared to independent single-object planning, and this improvement is maintained when scaling to three objects. Real-world experiments with two objects confirm a 29% reduction; the remaining gap to the theoretical 50% is attributed to inter-throw transition overhead. When target positions are randomly changed mid-execution, the system re-plans and successfully reaches the new targets within 100 ms latency without stopping the robot. These results establish the first unified planning framework for MOMT throwing -- demonstrating scalability to multiple objects in simulation and real-world feasibility on two-object tasks -- advancing the frontier of high-throughput robotic manipulation. A video summarizing the method and the hardware experiments is available at https://liuyangdh.github.io/momt-video

cs.RO↗

arXiv2026

Nonthermal nonlinear many-body excitonic dynamics in high-purity silicon

We report time-resolved optical pump-terahertz third-harmonic generation spectroscopy of picosecond nonthermal many-body charge dynamics in high-purity silicon at room temperature. Although in thermal equilibrium excitons in silicon are unstable due to a lower binding energy than the thermal energy at room temperature, following pulsed non-resonant optical pumping we can identify the formation of transient excitonic states on the picosecond time scale in our high-resolution nonlinear spectroscopic measurements. In contrast to unbound electron-hole pairs, the nonthermal many-body excitonic dynamics feature a characteristic time-dependent evolution, a distinct dependence on charge density, and a nonperturbative nonlinear terahertz response.

cond-mat.str-el↗

arXiv2026

DynaTokens: Controlling Token Dynamics for Continual Video-Language Understanding

Continual VideoQA with multimodal LLMs remains challenging because sequential adaptation induces task interference, while storing task-specific prompts becomes impractical as task sequences grow. We introduce DynaTokens, a transformer-based token generator that dynamically produces fine-tuning tokens on demand, enabling task-adaptive prompt updates through shared generation weights. To mitigate forgetting, we introduce meta-learning-inspired regularisers that look ahead to avoid task-specific sharp update directions while anchoring the evolving generator to prior-task behaviours. We theoretically connect this objective to sharpness-aware optimisation, showing how it favours flatter cross-task minima and improves retention. DynaTokens combines gradient-free routing based on robust pretrained token and visual embeddings with lightweight auxiliary multimodal supervision, reducing router drift during continual adaptation. Across standard continual VideoQA benchmarks, DynaTokens achieves higher average accuracy and substantially lower forgetting than strong baselines. It also improves zero-shot generalisation and remains effective in longer domain-incremental sequences with extended task shifts. Finally, we introduce a challenging ImageQA->VideoQA protocol and show that DynaTokens enables robust cross-modal continual transfer.

cs.CV↗

arXiv2026

EG-VQA: Benchmarking Verifiable Video Question Answering with Grounded Temporal Evidence

Recent advances in Video Large Language Models (Video-LLMs) have yielded promising performance on Video Question Answering (VideoQA). Nevertheless, existing benchmarks are predominantly evaluated through answer correctness, while the faithfulness of predicted evidence supporting those answers remains insufficiently evaluated. This disconnect between answer generation and evidence verification motivates the construction of the Evidence-Grounded Video Question Answering Benchmark (EG-VQA), a large-scale open-ended benchmark in which each QA pair is annotated with temporally localized textual evidence, requiring models to jointly produce answers and verifiable evidence. EG-VQA comprises 2,067 videos and 11,838 QA pairs with fine-grained evidence annotations. To evaluate predicted evidence, Evidence-Grounded F1 (EG-F1) is introduced as a metric that jointly measures temporal alignment and semantic consistency between predicted and ground-truth evidence. Experiments reveal a substantial discrepancy between answer correctness and evidence faithfulness: even strong proprietary models can achieve high answer accuracy while exhibiting gaps in temporal-semantic grounding. To investigate evidence-aware learning under EG-VQA, we develop EG-Reasoner, an evidence-aware model trained with explicit evidence supervision. EG-Reasoner achieves strong evidence grounding performance among open-source models while maintaining competitive answer generation results compared with proprietary systems. These findings highlight the importance of explicitly modeling evidence grounding and suggest that evidence-aware supervision provides an effective direction toward more reliable and interpretable VideoQA systems.

cs.CV↗

arXiv2026

Granularity-Adaptive Credit Assignment for Long-Horizon LLM Agent Reinforcement Learning

Long-horizon language-model agents trained with reinforcement learning oftenreceive sparse outcome rewards that do not reveal which decisions along a tra-jectory deserve credit. Episode-level advantages provide coarse trajectory-widecredit, while step-level comparisons offer finer resolution with context-dependentestimation noise. We propose Granularity-Adaptive Credit Assignment (GACA),a critic-free method that adaptively mixes episode- and step-level credit for eachdecision during policy optimization. GACA normalizes the sampled response'smean per-token negative log-likelihood (NLL) within each trajectory and uses theresulting criticality score to determine the step-specific mixture. The computationreuses rollout log-probabilities without additional training rollouts or model eval-uations. Our analysis characterizes optimal score-dependent mixing and derivesconditions linking expected NLL to a lower bound on the preferred step-levelweight. Across ALFWorld and WebShop with 1.5B and 7B backbones, GACAachieves the highest reported mean success rates among the compared methods,while introducing negligible additional computation.

cs.LG↗