Pengcheng Wang

Shizhe Diao, Pengcheng Wang, Yong Lin et al.

The increasing scale of large language models (LLMs) brings emergent abilities to various complex tasks requiring reasoning, such as arithmetic and commonsense reasoning. It is known that the effective design of task-specific prompts is critical for LLMs' ability to produce high-quality answers. In particular, an effective approach for complex question-and-answer tasks is example-based prompting with chain-of-thought (CoT) reasoning, which significantly improves the performance of LLMs. However, current CoT methods rely on a fixed set of human-annotated exemplars, which are not necessarily the most effective examples for different tasks. This paper proposes a new method, Active-Prompt, to adapt LLMs to different tasks with task-specific example prompts (annotated with human-designed CoT reasoning). For this purpose, we propose a solution to the key problem of determining which questions are the most important and helpful ones to annotate from a pool of task-specific queries. By borrowing ideas from the related problem of uncertainty-based active learning, we introduce several metrics to characterize the uncertainty so as to select the most uncertain questions for annotation. Experimental results demonstrate the superiority of our proposed method, achieving state-of-the-art on eight complex reasoning tasks. Further analyses of different uncertainty metrics, pool sizes, zero-shot learning, and accuracy-uncertainty relationship demonstrate the effectiveness of our method. Our code will be available at https://github.com/shizhediao/active-prompt.

Xuri Ge, Joemon M. Jose, Pengcheng Wang et al.

Facial palsy is unilateral facial nerve weakness or paralysis of rapid onset with unknown causes. Automatically estimating facial palsy severeness can be helpful for the diagnosis and treatment of people suffering from it across the world. In this work, we develop and experiment with a novel model for estimating facial palsy severity. For this, an effective Facial Action Units (AU) detection technique is incorporated into our model, where AUs refer to a unique set of facial muscle movements used to describe almost every anatomically possible facial expression. In this paper, we propose a novel Adaptive Local-Global Relational Network (ALGRNet) for facial AU detection and use it to classify facial paralysis severity. ALGRNet mainly consists of three main novel structures: (i) an adaptive region learning module that learns the adaptive muscle regions based on the detected landmarks; (ii) a skip-BiLSTM that models the latent relationships among local AUs; and (iii) a feature fusion&refining module that investigates the complementary between the local and global face. Quantitative results on two AU benchmarks, i.e., BP4D and DISFA, demonstrate our ALGRNet can achieve promising AU detection accuracy. We further demonstrate the effectiveness of its application to facial paralysis estimation by migrating ALGRNet to a facial paralysis dataset collected and annotated by medical professionals.

Liutong Han, Zhiyuan Tan, Hongbin Zhang et al.

The use of intrinsic functions to leverage hardware-specific capabilities is a crucial approach for optimizing library performance. Many mainstream libraries implement a large number of vectorized algorithms on Arm or x86 SIMD (Single-Instruction, Multiple-Data) intrinsic functions. Translating existing vectorized intrinsic code into the intrinsics of an emerging architecture is a practical and effective approach. However, current cross-architecture translation largely relies on manual rewriting or rule-based mapping methods, which are both time-consuming and prone to errors. We present \texttt{IntrinTrans}, a LLM-based agent that utilizes compile-and-test feedback to translate intrinsic code across architectures automatically, and further optimizes the generated intrinsics using register-usage information derived from liveness analysis. To evaluate the effectiveness of our method, we used \texttt{IntrinTrans} to translate the open-source benchmark from Arm Neon Intrinsic to the emerging RISC-V Vector (RVV) Intrinsic implementation and compared its performance with that of the native RVV implementation. Our experiments show that advanced LLMs can generate semantically correct RVV Intrinsic functions with only a finite number of iterations. Depending on the base LLMs, the pass rate ranges from 47% to 100%, achieving performance similar to the native implementation (0.85x to 1.28x).

Jiaxi Yin, Pengcheng Wang, Han Ding et al.

Accurate food intake detection is vital for dietary monitoring and chronic disease prevention. Traditional self-report methods are prone to recall bias, while camera-based approaches raise concerns about privacy. Furthermore, existing wearable-based methods primarily focus on a limited number of food types, such as hamburgers and pizza, failing to address the vast diversity of Chinese cuisine. To bridge this gap, we propose CuisineSense, a system that classifies Chinese food types by integrating hand motion cues from a smartwatch with head dynamics from smart glasses. To filter out irrelevant daily activities, we design a two-stage detection pipeline. The first stage identifies eating states by distinguishing characteristic temporal patterns from non-eating behaviors. The second stage then conducts fine-grained food type recognition based on the motions captured during food intake. To evaluate CuisineSense, we construct a dataset comprising 27.5 hours of IMU recordings across 11 food categories and 10 participants. Experiments demonstrate that CuisineSense achieves high accuracy in both eating state detection and food classification, offering a practical solution for unobtrusive, wearable-based dietary monitoring.The system code is publicly available at https://github.com/joeeeeyin/CuisineSense.git.

Pengcheng Wang, Kaiwen Hong, Chensheng Peng et al.

Unlike chatbots, physical AI must act while the world keeps evolving. Therefore, the inter-chunk pause of synchronous executors are fatal for dynamic tasks regardless of how fast the inference is. Asynchronous execution -- thinking while acting -- is therefore a structural requirement, and real-time chunking (RTC) makes it viable by recasting chunk transitions as inpainting: freezing committed actions and consistently generating the remainder. However, RTC with flow-matching policy is structurally suboptimal: its inpainting comes from inference-time corrections rather than the base policy, yielding little pre-training benefit, specific fine-tuning, heuristic guidance, and extra computation that inflates the latency. In this work, we observe that discrete diffusion policies, which generate actions by iteratively unmasking, are natural asynchronous executors that resolve all limitations at once: they are fine-tuning free since inpainting is their native operation, while early stopping further provides adaptive guidance and reduces inference cost. We propose DiscreteRTC, which replaces external corrections with native unmasking, and show on dynamic simulated benchmarks and real-world dynamic manipulation tasks that it achieves higher success rates than continuous RTC and other baselines. In summary, DiscreteRTC is simpler to implement with 0 lines of code for async inpainting, faster at inference with only 0.7x computation compared with generating actions from scratch, and better at execution with 50% higher success rate in real-world dynamic pick task compared with flow-matching-based RTC. More visualizations are on https://outsider86.github.io/DiscreteRTCSite/.

Pengcheng Wang, Jerry Huang, Jiarui Yao et al.

Language-model agent systems commonly rely on reactive prompting, in which a single instruction guides the model through an open-ended sequence of reasoning and tool-use steps, leaving control flow and intermediate state implicit and making agent behavior potentially difficult to control. Orchestration frameworks such as LangGraph, DSPy, and CrewAI impose greater structure through explicit workflow definitions, but tightly couple workflow logic with Python, making agents difficult to maintain and modify. In this paper, we introduce AgentSPEX, an Agent SPecification and EXecution Language for specifying LLM-agent workflows with explicit control flow and modular structure, along with a customizable agent harness. AgentSPEX supports typed steps, branching and loops, parallel execution, reusable submodules, and explicit state management, and these workflows execute within an agent harness that provides tool access, a sandboxed virtual environment, and support for checkpointing, verification, and logging. Furthermore, we provide a visual editor with synchronized graph and workflow views for authoring and inspection. We include ready-to-use agents for deep research and scientific research, and we evaluate AgentSPEX on 7 benchmarks. Finally, we show through a user study that AgentSPEX provides a more interpretable and accessible workflow-authoring paradigm than a popular existing agent framework.

Xiang Ma, Taihua Chen, Pengcheng Wang et al.

Time series forecasting is crucial for applications in various domains. Conventional methods often rely on global decomposition into trend, seasonal, and residual components, which become ineffective for real-world series dominated by local, complex, and highly dynamic patterns. Moreover, the high model complexity of such approaches limits their applicability in real-time or resource-constrained environments. In this work, we propose a novel \textbf{RE}liability-aware \textbf{C}odebook-\textbf{AS}sisted \textbf{T}ime series forecasting framework (\textbf{ReCast}) that enables lightweight and robust prediction by exploiting recurring local shapes. ReCast encodes local patterns into discrete embeddings through patch-wise quantization using a learnable codebook, thereby compactly capturing stable regular structures. To compensate for residual variations not preserved by quantization, ReCast employs a dual-path architecture comprising a quantization path for efficient modeling of regular structures and a residual path for reconstructing irregular fluctuations. A central contribution of ReCast is a reliability-aware codebook update strategy, which incrementally refines the codebook via weighted corrections. These correction weights are derived by fusing multiple reliability factors from complementary perspectives by a distributionally robust optimization (DRO) scheme, ensuring adaptability to non-stationarity and robustness to distribution shifts. Extensive experiments demonstrate that ReCast outperforms state-of-the-art (SOTA) models in accuracy, efficiency, and adaptability to distribution shifts.

Pengcheng Wang, Haoxiang Liu, Yang Dai et al.

CAPTCHAs are widely deployed as human verification mechanisms and frequently block intelligent agents from completing end-to-end automation in real-world web environments. Solving modern CAPTCHAs requires robust multi-step visual reasoning and interaction capabilities, yet training-based approaches have remained absent due to the lack of large-scale training data and process-level annotations. We introduce CaptchaBench, the first CAPTCHA benchmark designed to support large-scale training, comprising 16,000 programmatically generated samples across eight task categories with detailed region and process-level annotations. Systematic evaluation on CaptchaBench reveals that existing methods fail consistently on tasks requiring fine-grained visual detail capture and region-level comparison. We therefore present CaptchaMind, an RL-based solver trained with explicit reasoning process supervision, achieving 82.9% average success rate across eight tasks and 71.0% on real-world instances, substantially outperforming all existing methods without closed-source APIs.

Chiheng Lou, Sheng Qi, Rui Kang et al.

Deploying multiple models within shared GPU clusters is promising for improving resource efficiency in large language model (LLM) serving. Existing multi-LLM serving systems optimize GPU utilization at the cost of worse inference performance, especially time-to-first-token (TTFT). We identify the root cause of such compromise as their unawareness of future workload characteristics. In contrast, recent analysis on real-world traces has shown the high periodicity and long-term predictability of LLM serving workloads. We propose universal GPU workers to enable one-for-many GPU prewarming that loads models with knowledge of future workloads. Based on universal GPU workers, we design and build WarmServe, a multi-LLM serving system that (1) mitigates cluster-wide prewarming interference by adopting an evict-aware model placement strategy, (2) prepares universal GPU workers in advance by proactive prewarming, and (3) manages GPU memory with a zero-overhead memory switching mechanism. Evaluation under real-world datasets shows that WarmServe improves TTFT by up to 50.8$\times$ compared to the state-of-the-art autoscaling-based system, while being capable of serving up to 2.5$\times$ more requests compared to the GPU-sharing system.

Jiawei Yi, Ping Gong, Youhui Bai et al.

The growth of million-token LLMs exposes the scalability limits of inference systems, where the KVCache dominates memory usage and data transfer overhead. Recent offloading systems migrate the KVCache to CPU memory and incorporate top-k attention to reduce the volume of data transferred from the CPU, while further applying system-level optimizations such as on-GPU caching and prefetching to lower transfer overhead. However, they overlook the CPU bottleneck in three aspects: (1) substantial overhead of fine-grained dynamic cache management performed on the CPU side, (2) significant transfer overhead from poor PCIe bandwidth utilization caused by heavy gathering operations at the CPU side, and (3) GPU runtime bubbles introduced by coarse-grained CPU-centric synchronization. To address these challenges, we propose CLO, a CPU-light KVCache offloading system via algorithm-system co-design. CLO features: (1) a coarse-grained head-wise approximate on-GPU caching strategy with negligible cache management cost, (2) seamless combination of data prefetching and on-GPU persistent caching for lower transfer overhead, (3) a zero-copy transfer engine to fully exploit PCIe bandwidth, and a GPU-centric synchronization method to eliminate GPU stalls. Evaluation on two widely-used LLMs demonstrates that CLO achieves comparable accuracy to state-of-the-art systems, while substantially minimizing CPU overhead, fully utilizing PCIe bandwidth, thus improving decoding throughput by 9.3%-66.6%. Our results highlight that algorithm-system co-design is essential for memory-constrained LLM inference on modern GPU platforms. We open source CLO at https://github.com/CommediaJW/CLO.

Jianglan Wei, Zhenyu Zhang, Pengcheng Wang et al.

Energy-efficient medical data classification is essential for modern disease screening, particularly in home and field healthcare where embedded devices are prevalent. While deep learning models achieve state-of-the-art accuracy, their substantial energy consumption and reliance on GPUs limit deployment on such platforms. We present HDC-X, a lightweight classification framework designed for low-power devices. HDC-X encodes data into high-dimensional hypervectors, aggregates them into multiple cluster-specific prototypes, and performs classification through similarity search in hyperspace. We evaluate HDC-X across three medical classification tasks; on heart sound classification, HDC-X is $350\times$ more energy-efficient than Bayesian ResNet with less than 1% accuracy difference. Moreover, HDC-X demonstrates exceptional robustness to noise, limited training data, and hardware error, supported by both theoretical analysis and empirical results, highlighting its potential for reliable deployment in real-world settings. Code is available at https://github.com/jianglanwei/HDC-X.

Fei Wang, Tingting Zhang, Xilei Wu et al.

Hand hygiene is among the most effective daily practices for preventing infectious diseases such as influenza, malaria, and skin infections. While professional guidelines emphasize proper handwashing to reduce the risk of viral infections, surveys reveal that adherence to these recommendations remains low. To address this gap, we propose UWash, a wearable solution leveraging smartwatches to evaluate handwashing procedures, aiming to raise awareness and cultivate high-quality handwashing habits. We frame the task of handwashing assessment as an action segmentation problem, similar to those in computer vision, and introduce a simple yet efficient two-stream UNet-like network to achieve this goal. Experiments involving 51 subjects demonstrate that UWash achieves 92.27% accuracy in handwashing gesture recognition, an error of <0.5 seconds in onset/offset detection, and an error of <5 points in gesture scoring under user-dependent settings. The system also performs robustly in user-independent and user-independent-location-independent evaluations. Remarkably, UWash maintains high performance in real-world tests, including evaluations with 10 random passersby at a hospital 9 months later and 10 passersby in an in-the-wild test conducted 2 years later. UWash is the first system to score handwashing quality based on gesture sequences, offering actionable guidance for improving daily hand hygiene. The code and dataset are publicly available at https://github.com/aiotgroup/UWash

Karthick Shankar, Pengcheng Wang, Ran Xu et al.

With diverse IoT workloads, placing compute and analytics close to where data is collected is becoming increasingly important. We seek to understand what is the performance and the cost implication of running analytics on IoT data at the various available platforms. These workloads can be compute-light, such as outlier detection on sensor data, or compute-intensive, such as object detection from video feeds obtained from drones. In our paper, JANUS, we profile the performance/$ and the compute versus communication cost for a compute-light IoT workload and a compute-intensive IoT workload. In addition, we also look at the pros and cons of some of the proprietary deep-learning object detection packages, such as Amazon Rekognition, Google Vision, and Azure Cognitive Services, to contrast with open-source and tunable solutions, such as Faster R-CNN (FRCNN). We find that AWS IoT Greengrass delivers at least 2X lower latency and 1.25X lower cost compared to all other cloud platforms for the compute-light outlier detection workload. For the compute-intensive streaming video analytics task, an opensource solution to object detection running on cloud VMs saves on dollar costs compared to proprietary solutions provided by Amazon, Microsoft, and Google, but loses out on latency (up to 6X). If it runs on a low-powered edge device, the latency is up to 49X lower.

Dayiheng Liu, Yu Yan, Yeyun Gong et al.

Multi-task benchmarks such as GLUE and SuperGLUE have driven great progress of pretraining and transfer learning in Natural Language Processing (NLP). These benchmarks mostly focus on a range of Natural Language Understanding (NLU) tasks, without considering the Natural Language Generation (NLG) models. In this paper, we present the General Language Generation Evaluation (GLGE), a new multi-task benchmark for evaluating the generalization capabilities of NLG models across eight language generation tasks. For each task, we continue to design three subtasks in terms of task difficulty (GLGE-Easy, GLGE-Medium, and GLGE-Hard). This introduces 24 subtasks to comprehensively compare model performance. To encourage research on pretraining and transfer learning on NLG models, we make GLGE publicly available and build a leaderboard with strong baselines including MASS, BART, and ProphetNet (The source code and dataset are publicly available at https://github.com/microsoft/glge).

Ziran Liu, Wei Wang, Jinhao Wang et al.

This paper develops the angular and static-channel component of Geometric and Spectral Alignment for residual Jacobian chains. Starting from Cartan-coordinate rigidity and fitted effective-rank windows, we study how dominant singular subspaces are transported across adjacent layers and how the resulting finite matrices can be displayed in physical channel coordinates. The main results are deterministic, margin-verified results. We bound the error between full interface transport and its dominant-window truncation, add fitted-tail errors so that empirical spectra can be certified against the Gibbs--Cartan tail model, and distinguish source-mode incidence from fully physical input-output channel incidence. Given row groups and active supports, the Physical Alignment Matrix decomposes orthogonally as core plus overlap plus noise. Active-column gaps, pairwise overlap margins, and noise bounds combine into a static certificate radius under which the full transport and the truncated transport induce the same active supports, pairwise incidence graph, SRS sets, hub columns, and core/overlap/noise masks. The finer SC/SA/ST labels of the Invariant Channel Mapping require additional row-energy and profile-correlation margins, stated as explicit perturbation tests. The empirical section reports the matrices and block-energy heatmaps that measure these certificate quantities across CNNs, language models, and vision/diffusion backbones. The figures are interpreted as finite-dimensional measurements; complete membership in the Physical GSA certificate domain requires checking the numerical margin protocol stated in Section 10.

Ziran Liu, Wei Wang, Jinhao Wang et al.

Deep residual architectures are modeled as products of near-identity Jacobians. This paper proves deterministic quotient-geometric estimates for singular spectra of Frobenius-normalized layer factors, emphasizing a normalized top-radial Cartan coordinate and fitted power-law chart. Full-rank factors are mapped from $\mathrm{GL}(d)$ to the positive cone by $A\mapsto A^\top A$, then to ordered eigenvalue data. Under Frobenius normalization, exact power-law spectra form a trace-normalized Cartan orbit. This orbit is a Gibbs family on ranks, a Fisher information line, and a Bures--Wasserstein curve with line element $d/4$ times Fisher information. The main rigidity theorem is a slack-aware margin inequality: interface radial amplitude, non-backtracking slack, and signed residual variation control displacement of the fitted Cartan coordinate. In the exact-chart zero-slack case, a depth-$L$ budget gives exponent drift of order $(\log M)/L$; generally, slack and residual increments augment the bound. We separate scalar top-radial from full-Cartan spectral control, which also needs Bures/Hellinger residual variation. We prove approximate-power-law and metric-chart versions, converse lower bounds, Fisher--KL/Bures action estimates, and near-identity expansions for normalized residual chains. Near-identity results verify transport budgets; chart quality remains measurable. Effective rank is a spectral-energy quantile, giving finite-width power-law tail bounds and robust rank-window transition estimates. Empirical static-weight exponent profiles serve as diagnostics; full verification also requires interface budgets, slacks, and residuals for the same operator chain.

Hanning Zhang, Pengcheng Wang, Shizhe Diao et al.

Large language models (LLMs) have shown promise in performing complex multi-step reasoning, yet they continue to struggle with mathematical reasoning, often making systematic errors. A promising solution is reinforcement learning (RL) guided by reward models, particularly those focusing on process rewards, which score each intermediate step rather than solely evaluating the final outcome. This approach is more effective at guiding policy models towards correct reasoning trajectories. In this work, we propose an entropy-regularized process reward model (ER-PRM) that integrates KL-regularized Markov Decision Processes (MDP) to balance policy optimization with the need to prevent the policy from shifting too far from its initial distribution. We derive a novel reward construction method based on the theoretical results. Our theoretical analysis shows that we could derive the optimal reward model from the initial policy sampling. Our empirical experiments on the MATH and GSM8K benchmarks demonstrate that ER-PRM consistently outperforms existing process reward models, achieving 1% improvement on GSM8K and 2-3% improvement on MATH under best-of-N evaluation, and more than 1% improvement under RLHF. These results highlight the efficacy of entropy-regularization in enhancing LLMs' reasoning capabilities.

Changjian Chen, Fei Lv, Yalong Guan et al.

The performance of computer vision models in certain real-world applications (e.g., rare wildlife observation) is limited by the small number of available images. Expanding datasets using pre-trained generative models is an effective way to address this limitation. However, since the automatic generation process is uncontrollable, the generated images are usually limited in diversity, and some of them are undesired. In this paper, we propose a human-guided image generation method for more controllable dataset expansion. We develop a multi-modal projection method with theoretical guarantees to facilitate the exploration of both the original and generated images. Based on the exploration, users refine the prompts and re-generate images for better performance. Since directly refining the prompts is challenging for novice users, we develop a sample-level prompt refinement method to make it easier. With this method, users only need to provide sample-level feedback (e.g., which samples are undesired) to obtain better prompts. The effectiveness of our method is demonstrated through the quantitative evaluation of the multi-modal projection method, improved model performance in the case study for both classification and object detection tasks, and positive feedback from the experts.

Haotian Lin, Pengcheng Wang, Jeff Schneider et al.

Model-based reinforcement learning algorithms that combine model-based planning and learned value/policy prior have gained significant recognition for their high data efficiency and superior performance in continuous control. However, we discover that existing methods that rely on standard SAC-style policy iteration for value learning, directly using data generated by the planner, often result in \emph{persistent value overestimation}. Through theoretical analysis and experiments, we argue that this issue is deeply rooted in the structural policy mismatch between the data generation policy that is always bootstrapped by the planner and the learned policy prior. To mitigate such a mismatch in a minimalist way, we propose a policy regularization term reducing out-of-distribution (OOD) queries, thereby improving value learning. Our method involves minimum changes on top of existing frameworks and requires no additional computation. Extensive experiments demonstrate that the proposed approach improves performance over baselines such as TD-MPC2 by large margins, particularly in 61-DoF humanoid tasks. View qualitative results at https://darthutopian.github.io/tdmpc_square/.

Weihua Gao, Wenlong Niu, Wenlong Lu et al.

The detection and tracking of small targets in passive optical remote sensing (PORS) has broad applications. However, most of the previously proposed methods seldom utilize the abundant temporal features formed by target motion, resulting in poor detection and tracking performance for low signal-to-clutter ratio (SCR) targets. In this article, we analyze the difficulty based on spatial features and the feasibility based on temporal features of realizing effective detection. According to this analysis, we use a multi-frame as a detection unit and propose a detection method based on temporal energy selective scaling (TESS). Specifically, we investigated the composition of intensity temporal profiles (ITPs) formed by pixels on a multi-frame detection unit. For the target-present pixel, the target passing through the pixel will bring a weak transient disturbance on the ITP and introduce a change in the statistical properties of ITP. We use a well-designed function to amplify the transient disturbance, suppress the background and noise components, and output the trajectory of the target on the multi-frame detection unit. Subsequently, to solve the contradiction between the detection rate and the false alarm rate brought by the traditional threshold segmentation, we associate the temporal and spatial features of the output trajectory and propose a trajectory extraction method based on the 3D Hough transform. Finally, we model the trajectory of the target and propose a trajectory-based multi-target tracking method. Compared with the various state-of-the-art detection and tracking methods, experiments in multiple scenarios prove the superiority of our proposed methods.

Pengcheng Wang, Sheng Li, Takahiro Shinozaki

In this paper, we propose RAG-Boost (ST-ShinozakiLab Task I system), which enhances the baseline LLM-based ASR system of the MLC-SLM Challenge (task I) with a retrieval-augmented generation (RAG) module on the fly. Each partial ASR hypothesis queries a vector store of audio-text pairs and domain terms, and the retrieved results are fused with the live ASR hypotheses to fix recognition errors. The fused hypotheses are passed to the LLM, yielding improved responses.

Pengcheng Wang, Xinghao Zhu, Yuxin Chen et al.

Reinforcement Learning and Imitation Learning have achieved widespread success in many domains but remain constrained during real-world deployment. One of the main issues is the additional requirements that were not considered during training. To address this challenge, policy customization has been introduced, aiming to adapt a prior policy while preserving its inherent properties and meeting new task-specific requirements. A principled approach to policy customization is Residual Q-Learning (RQL), which formulates the problem as a Markov Decision Process (MDP) and derives a family of value-based learning algorithms. However, RQL has not yet been applied to policy gradient methods, which restricts its applicability, especially in tasks where policy gradient has already proven more effective. In this work, we first derive a concise form of Soft Policy Gradient as a preliminary. Building on this, we introduce Residual Policy Gradient (RPG), which extends RQL to policy gradient methods, allowing policy customization in gradient-based RL settings. With the view of RPG, we rethink the KL-regularized objective widely used in RL fine-tuning. We show that under certain assumptions, KL-regularized objective leads to a maximum-entropy policy that balances the inherent properties and task-specific requirements on a reward-level. Our experiments in MuJoCo demonstrate the effectiveness of Soft Policy Gradient and Residual Policy Gradient.

Pengcheng Wang, Qinghang Liu, Haotian Lin et al.

Learning domain adaptive policies that can generalize to unseen transition dynamics, remains a fundamental challenge in learning-based control. Substantial progress has been made through domain representation learning to capture domain-specific information, thus enabling domain-aware decision making. We analyze the process of learning domain representations through dynamical prediction and find that selecting contexts adjacent to the current step causes the learned representations to entangle static domain information with varying dynamical properties. Such mixture can confuse the conditioned policy, thereby constraining zero-shot adaptation. To tackle the challenge, we propose DADP (Domain Adaptive Diffusion Policy), which achieves robust adaptation through unsupervised disentanglement and domain-aware diffusion injection. First, we introduce Lagged Context Dynamical Prediction, a strategy that conditions future state estimation on a historical offset context; by increasing this temporal gap, we unsupervisedly disentangle static domain representations by filtering out transient properties. Second, we integrate the learned domain representations directly into the generative process by biasing the prior distribution and reformulating the diffusion target. Extensive experiments on challenging benchmarks across locomotion and manipulation demonstrate the superior performance, and the generalizability of DADP over prior methods. More visualization results are available on the https://outsider86.github.io/DomainAdaptiveDiffusionPolicy/.

Guojian Zhan, Likun Wang, Pengcheng Wang et al.

Maximum entropy has become a mainstream off-policy reinforcement learning (RL) framework for balancing exploitation and exploration. However, two bottlenecks still limit further performance improvement: (1) non-stationary Q-value estimation caused by jointly injecting entropy and updating its weighting parameter, i.e., temperature; and (2) short-sighted local entropy tuning that adjusts temperature only according to the current single-step entropy, without considering the effect of cumulative entropy over time. In this paper, we extends maximum entropy framework by proposing a trajectory entropy-constrained reinforcement learning (TECRL) framework to address these two challenges. Within this framework, we first separately learn two Q-functions, one associated with reward and the other with entropy, ensuring clean and stable value targets unaffected by temperature updates. Then, the dedicated entropy Q-function, explicitly quantifying the expected cumulative entropy, enables us to enforce a trajectory entropy constraint and consequently control the policy long-term stochasticity. Building on this TECRL framework, we develop a practical off-policy algorithm, DSAC-E, by extending the state-of-the-art distributional soft actor-critic with three refinements (DSAC-T). Empirical results on the OpenAI Gym benchmark demonstrate that our DSAC-E can achieve higher returns and better stability.

Changjian Chen, Pengcheng Wang, Fei Lyu et al.

Hybrid rice breeding crossbreeds different rice lines and cultivates the resulting hybrids in fields to select those with desirable agronomic traits, such as higher yields. Recently, genomic selection has emerged as an efficient way for hybrid rice breeding. It predicts the traits of hybrids based on their genes, which helps exclude many undesired hybrids, largely reducing the workload of field cultivation. However, due to the limited accuracy of genomic prediction models, breeders still need to combine their experience with the models to identify regulatory genes that control traits and select hybrids, which remains a time-consuming process. To ease this process, in this paper, we proposed a visual analysis method to facilitate interactive hybrid rice breeding. Regulatory gene identification and hybrid selection naturally ensemble a dual-analysis task. Therefore, we developed a parametric dual projection method with theoretical guarantees to facilitate interactive dual analysis. Based on this dual projection method, we further developed a gene visualization and a hybrid visualization to verify the identified regulatory genes and hybrids. The effectiveness of our method is demonstrated through the quantitative evaluation of the parametric dual projection method, identified regulatory genes and desired hybrids in the case study, and positive feedback from breeders.

James Wen, Sahil Nalawade, Zhiwei Liang et al. · deepmind, harvard

Background: Generative artificial intelligence (AI) deployment in academic medical settings raises copyright compliance concerns. Dana-Farber Cancer Institute implemented GPT4DFCI, an internal generative AI tool utilizing OpenAI models, that is approved for enterprise use in research and operations. Given (1) the exceptionally broad adoption of the tool in our organization, (2) our research mission, and (3) the shared responsibility model required to benefit from Customer Copyright Commitment in Azure OpenAI Service products, we deemed rigorous copyright compliance testing necessary. Case Description: We conducted a structured red teaming exercise in Nov. 2024, with 42 participants from academic, industry, and government institutions. Four teams attempted to extract copyrighted content from GPT4DFCI across four domains: literary works, news articles, scientific publications, and access-restricted clinical notes. Teams successfully extracted verbatim book dedications and near-exact passages through various strategies. News article extraction failed despite jailbreak attempts. Scientific article reproduction yielded only high-level summaries. Clinical note testing revealed appropriate privacy safeguards. Discussion: The successful extraction of literary content indicates potential copyrighted material presence in training data, necessitating inference-time filtering. Differential success rates across content types suggest varying protective mechanisms. The event led to implementation of a copyright-specific meta-prompt in GPT4DFCI; this mitigation has been in production since Jan. 2025. Conclusion: Systematic red teaming revealed specific vulnerabilities in generative AI copyright compliance, leading to concrete mitigation strategies. Academic medical institutions deploying generative AI should implement continuous testing protocols to ensure legal and ethical compliance.

Jayoung Lee, PengCheng Wang, Ran Xu et al.

Efficient and adaptive computer vision systems have been proposed to make computer vision tasks, such as image classification and object detection, optimized for embedded or mobile devices. These solutions, quite recent in their origin, focus on optimizing the model (a deep neural network, DNN) or the system by designing an adaptive system with approximation knobs. In spite of several recent efforts, we show that existing solutions suffer from two major drawbacks. First, the system does not consider energy consumption of the models while making a decision on which model to run. Second, the evaluation does not consider the practical scenario of contention on the device, due to other co-resident workloads. In this work, we propose an efficient and adaptive video object detection system, Virtuoso, which is jointly optimized for accuracy, energy efficiency, and latency. Underlying Virtuoso is a multi-branch execution kernel that is capable of running at different operating points in the accuracy-energy-latency axes, and a lightweight runtime scheduler to select the best fit execution branch to satisfy the user requirement. To fairly compare with Virtuoso, we benchmark 15 state-of-the-art or widely used protocols, including Faster R-CNN (FRCNN), YOLO v3, SSD, EfficientDet, SELSA, MEGA, REPP, FastAdapt, and our in-house adaptive variants of FRCNN+, YOLO+, SSD+, and EfficientDet+ (our variants have enhanced efficiency for mobiles). With this comprehensive benchmark, Virtuoso has shown superiority to all the above protocols, leading the accuracy frontier at every efficiency level on NVIDIA Jetson mobile GPUs. Specifically, Virtuoso has achieved an accuracy of 63.9%, which is more than 10% higher than some of the popular object detection models, FRCNN at 51.1%, and YOLO at 49.5%.

Pengcheng Wang, Lingqiao Ji, Zhilong Ji et al.

In this technical report, we briefly introduce the solution of our team "TAL-ai" for (Semi-) supervised Face detection in the low light condition in UG2+ Challenge in CVPR 2021. By conducting several experiments with popular image enhancement methods and image transfer methods, we pulled the low light image and the normal image to a more closer domain. And it is observed that using these data to training can achieve better performance. We also adapt several popular object detection frameworks, e.g., DetectoRS, Cascade-RCNN, and large backbone like Swin-transformer. Finally, we ensemble several models which achieved mAP 74.89 on the testing set, ranking 1st on the final leaderboard.

Ran Xu, Chen-lin Zhang, Pengcheng Wang et al.

Advanced video analytic systems, including scene classification and object detection, have seen widespread success in various domains such as smart cities and autonomous transportation. With an ever-growing number of powerful client devices, there is incentive to move these heavy video analytics workloads from the cloud to mobile devices to achieve low latency and real-time processing and to preserve user privacy. However, most video analytic systems are heavyweight and are trained offline with some pre-defined latency or accuracy requirements. This makes them unable to adapt at runtime in the face of three types of dynamism -- the input video characteristics change, the amount of compute resources available on the node changes due to co-located applications, and the user's latency-accuracy requirements change. In this paper we introduce ApproxDet, an adaptive video object detection framework for mobile devices to meet accuracy-latency requirements in the face of changing content and resource contention scenarios. To achieve this, we introduce a multi-branch object detection kernel (layered on Faster R-CNN), which incorporates a data-driven modeling approach on the performance metrics, and a latency SLA-driven scheduler to pick the best execution branch at runtime. We couple this kernel with approximable video object tracking algorithms to create an end-to-end video object detection system. We evaluate ApproxDet on a large benchmark video dataset and compare quantitatively to AdaScale and YOLOv3. We find that ApproxDet is able to adapt to a wide variety of contention and content characteristics and outshines all baselines, e.g., it achieves 52% lower latency and 11.1% higher accuracy over YOLOv3.

Pengcheng Wang, Zihao Wang, Zhilong Ji et al.

This paper introduces our approach to the EmotioNet Challenge 2020. We pose the AU recognition problem as a multi-task learning problem, where the non-rigid facial muscle motion (mainly the first 17 AUs) and the rigid head motion (the last 6 AUs) are modeled separately. The co-occurrence of the expression features and the head pose features are explored. We observe that different AUs converge at various speed. By choosing the optimal checkpoint for each AU, the recognition results are improved. We are able to obtain a final score of 0.746 in validation set and 0.7306 in the test set of the challenge.

Ming Wu, Pengcheng Wang, Kangqi Yin et al.

To detect large-variance code clones (i.e. clones with relatively more differences) in large-scale code repositories is difficult because most current tools can only detect almost identical or very similar clones. It will make promotion and changes to some software applications such as bug detection, code completion, software analysis, etc. Recently, CCAligner made an attempt to detect clones with relatively concentrated modifications called large-gap clones. Our contribution is to develop a novel and effective detection approach of large-variance clones to more general cases for not only the concentrated code modifications but also the scattered code modifications. A detector named LVMapper is proposed, borrowing and changing the approach of sequencing alignment in bioinformatics which can find two similar sequences with more differences. The ability of LVMapper was tested on both self-synthetic datasets and real cases, and the results show substantial improvement in detecting large-variance clones compared with other state-of-the-art tools including CCAligner. Furthermore, our new tool also presents good recall and precision for general Type-1, Type-2 and Type-3 clones on the widely used benchmarking dataset, BigCloneBench.

Ran Xu, Rakesh Kumar, Pengcheng Wang et al.

Videos take a lot of time to transport over the network, hence running analytics on the live video on embedded or mobile devices has become an important system driver. Considering that such devices, e.g., surveillance cameras or AR/VR gadgets, are resource constrained, creating lightweight deep neural networks (DNNs) for embedded devices is crucial. None of the current approximation techniques for object classification DNNs can adapt to changing runtime conditions, e.g., changes in resource availability on the device, the content characteristics, or requirements from the user. In this paper, we introduce ApproxNet, a video object classification system for embedded or mobile clients. It enables novel dynamic approximation techniques to achieve desired inference latency and accuracy trade-off under changing runtime conditions. It achieves this by enabling two approximation knobs within a single DNN model, rather than creating and maintaining an ensemble of models (e.g., MCDNN [MobiSys-16]. We show that ApproxNet can adapt seamlessly at runtime to these changes, provides low and stable latency for the image and video frame classification problems, and show the improvement in accuracy and latency over ResNet [CVPR-16], MCDNN [MobiSys-16], MobileNets [Google-17], NestDNN [MobiCom-18], and MSDNet [ICLR-18].