arXiv subjects

Qiwei Li

Publications and source records attributed to Qiwei Li.

At least 19 recordsLinked to original sources

arXiv2026

teLLMe Why (Ain't Nothing but a Jam): Exploratory Causal Analysis of Urban Driving Data

Traffic agencies now have access to large volumes of video-derived data for studying safety and congestion. Most of these data are observational and collected without interventions, which makes causal questions such as "How would rain change traffic density?" difficult to answer. We present teLLMe, a system for exploratory causal analysis of urban driving datasets. The system starts from a structured event table built from dashcam annotations and combines causal structure learning with the PC algorithm, bootstrap-based stability checks, and query-specific effect estimation using linear regression and DoWhy. Natural-language questions are mapped to structured causal queries through a schema-aware LLM, enabling users to specify treatments, outcomes, and subpopulations. teLLMe returns a "Causal Card" that summarizes effect estimates, adjustment sets, DAG support, and assumptions, followed by a short natural-language explanation. Case studies on BDD-derived traffic events show that the system can surface plausible relationships involving weather, peak hours, and traffic density, while making uncertainty and modeling choices explicit. The system is designed as a tool for hypothesis generation and expert reasoning rather than a source of definitive causal claims.

cs.AI↗

arXiv2026

Xiaomi-Robotics-U0: Unified Embodied Synthesis with World Foundation Model

Recent foundation image and video generation models offer strong generalization and controllability, but their direct application to embodied scenarios is limited by requirements for multi-view consistency, geometric coherence, and robot embodiment constraints. Existing methods typically adapt foundation models with limited robot data, often sacrificing visual knowledge acquired during large-scale pre-training. We present Xiaomi-Robotics-U0, a 38-billion-parameter multimodal autoregressive model for unified embodied synthesis. It treats embodied generation as an extension of foundation image and video generation and jointly optimizes text-to-image generation, image editing, embodied scene generation, embodied transfer, and embodied video generation. This unified framework preserves the generalization of the pre-trained world foundation model while adapting it to embodied settings. Xiaomi-Robotics-U0 is the first model to support high-quality multi-view scene generation across multiple robot embodiments and to introduce structured, controllable embodied transfer for fine-grained editing while preserving multi-view consistency and interaction dynamics. It achieves state-of-the-art results on single-step and sequential generation tasks, outperforming GPT-Image-2.0 in human evaluations of embodied scene generation and transfer, ranking first on World Arena for embodied video generation, and improving the out-of-distribution success rate of pi_0.5 from 36.9% to 63.2% on challenging real-world manipulation tasks. These results show that foundation world models can serve both as embodied world models and scalable data engines for embodied intelligence. Code and checkpoints are available at https://robotics.xiaomi.com/xiaomi-robotics-u0.html.

cs.RO↗

arXiv2026

Extending Embodied Question Answering from Perception to Decision

Embodied Question Answering (EQA) connects perception, reasoning, and interaction within embodied environments. However, existing datasets and benchmarks remain fragmented, each focusing on a limited subset of reasoning skills such as spatial understanding or procedural reasoning, without offering a unified large-scale framework for comprehensive evaluation. We present EQA-Decision, a large-scale embodied QA dataset that systematically covers four complementary dimensions of embodied reasoning: static scene construction, spatial understanding, task dynamics reasoning, and instant decision. The dataset contains over four million question-answer pairs with hierarchical annotations across diverse embodied scenarios. In addition, we develop RoboDecision, a strong baseline model aligned with the EQA-Decision Benchmark, providing a unified framework that jointly evaluates perception, reasoning, and action-level decision-making in embodied environments. Results demonstrate that EQA-Decision effectively benchmarks and enhances VLM capabilities in spatial and interaction reasoning, providing a solid foundation for advancing embodied intelligence research.

cs.RO↗

arXiv2026

RotVLA: Rotational Latent Action for Vision-Language-Action Model

Latent Action Models (LAMs) have emerged as an effective paradigm for handling heterogeneous datasets during Vision-Language-Action (VLA) model pretraining, offering a unified action space across embodiments. However, existing LAMs often rely on discrete quantization encode and decode pipelines, which can lead to trivial frame reconstruction behavior, limited representational capacity, and a lack of physically meaningful structure. We introduce RotVLA, a VLA framework built on a continuous rotational latent action representation. Latent actions are modeled as elements of SO(n), providing continuity, compositionality, and structured geometry aligned with real-world action dynamics. A triplet frame learning framework further enforces meaningful temporal dynamics while avoiding degeneration. RotVLA consists of a VLM backbone and a flow-matching action head, pretrained on large-scale cross-embodiment robotic datasets and human videos with latent-action supervision. For downstream robot control, the flow-matching head is extended into a unified action expert that jointly denoises latent and robot actions. Here, latent actions serve as a latent planner, providing high-level guidance that conditions action generation. With only 1.7B parameters and 1700+ hours of pretraining data, RotVLA achieves 98.2% on LIBERO and 89.6% / 88.5% on RoboTwin2.0 under clean and randomized settings, respectively. It also demonstrates strong real-world performance on manipulation tasks, consistently outperforming existing VLA models.

cs.RO↗

arXiv2026

TraceFix: Repairing Agent Coordination Protocols with TLA+ Counterexamples

We present TraceFix, a verification-first pipeline for Large Language Model (LLM) multi-agent coordination. An agent synthesizes a protocol topology as a structured intermediate representation (IR) from a task description, generates PlusCal coordination logic, and iteratively repairs the protocol using counterexamples from the TLA+ model checker (TLC) until verification succeeds. Verified process bodies are compiled into per-agent system prompts and executed under a runtime monitor that rejects out-of-topology coordination operations. On 48 tasks spanning 16 scenario families, all tasks reach full TLC verification; 62.5% pass on the first attempt and none requires more than four repair iterations. State spaces span six orders of magnitude yet verification completes in under 60 s for every task. A 3,456-run runtime comparison shows that topology-monitored execution achieves the highest task completion (89.4% average, 81.5% full) and that runtimes using the verified protocol degrade at roughly half the rate of prompt-only and chat-only baselines when model capability is reduced. A paired ablation under a fixed runtime shows that TLC-verified protocols cut deadlock/livelock (DL/LL) from 31.1% to 14.1%, with the largest separation under fault injection.

cs.AI↗

arXiv2026

Unified 4D World Action Modeling from Video Priors with Asynchronous Denoising

We propose X-WAM, a Unified 4D World Model that unifies real-time robotic action execution and high-fidelity 4D world synthesis (video + 3D reconstruction) in a single framework, addressing the critical limitations of prior unified world models (e.g., UWM) that only model 2D pixel-space and fail to balance action efficiency and world modeling quality. To leverage the strong visual priors of pretrained video diffusion models, X-WAM imagines the future world by predicting multi-view RGB-D videos, and obtains spatial information efficiently through a lightweight structural adaptation: replicating the final few blocks of the pretrained Diffusion Transformer into a dedicated depth prediction branch for the reconstruction of future spatial information. Moreover, we propose Asynchronous Noise Sampling (ANS) to jointly optimize generation quality and action decoding efficiency. ANS applies a specialized asynchronous denoising schedule during inference, which rapidly decodes actions with fewer steps to enable efficient real-time execution, while dedicating the full sequence of steps to generate high-fidelity video. Rather than entirely decoupling the timesteps during training, ANS samples from their joint distribution to align with the inference distribution. Pretrained on over 5,800 hours of robotic data, X-WAM achieves 79.2% and 90.7% average success rate on RoboCasa and RoboTwin 2.0 benchmarks, while producing high-fidelity 4D reconstruction and generation surpassing existing methods in both visual and geometric metrics.

cs.RO↗

arXiv2026

ProgressVLA: Progress-Guided Diffusion Policy for Vision-Language Robotic Manipulation

Most existing vision-language-action (VLA) models for robotic manipulation lack progress awareness, typically relying on hand-crafted heuristics for task termination. This limitation is particularly severe in long-horizon tasks involving cascaded sub-goals. In this work, we investigate the estimation and integration of task progress, proposing a novel model named {\textbf \vla}. Our technical contributions are twofold: (1) \emph{robust progress estimation}: We pre-train a progress estimator on large-scale, unsupervised video-text robotic datasets. This estimator achieves a low prediction residual (0.07 on a scale of $[0, 1]$) in simulation and demonstrates zero-shot generalization to unseen real-world samples, and (2) \emph{differentiable progress guidance}: We introduce an inverse dynamics world model that maps predicted action tokens into future latent visual states. These latents are then processed by the progress estimator; by applying a maximal progress regularization, we establish a differentiable pipeline that provides progress-piloted guidance to refine action tokens. Extensive experiments on the CALVIN and LIBERO benchmarks, alongside real-world robot deployment, consistently demonstrate substantial improvements in success rates and generalization over strong baselines.

cs.RO↗

arXiv2025

Spatial Analysis for AI-segmented Histopathology Images: Methods and Implementation

Quantitative characterization of cellular spatial organization is critical for understanding tumor progression and immune response. Recent advances in artificial intelligence (AI) enable large-scale segmentation and classification of nuclei from digitized histopathology slides, producing massive point pattern and marked point pattern data. However, accessible and standardized tools for downstream spatial statistical analysis remain limited. We present SASHIMI (Spatial Analysis for Segmented Histopathology Images using Machine Intelligence), a browser-based platform for real-time spatial analysis of AI-segmented histopathology images. Rather than proposing new spatial methods, SASHIMI systematically organizes and operationalizes 27 widely used spatial summary statistics, areal indices, and topological features within a unified computational framework. The platform computes mathematically grounded descriptors including K-, L-, G-, F-, and J-functions, pair correlation and mark connection functions, spatial autocorrelation measures, similarity indices, and persistent homology-based topological summaries. Outputs include both functional curves and scalar feature tables suitable for downstream statistical modeling. We illustrate the framework using two cancer cohorts: oral potentially malignant disorders and non-small-cell lung cancer. Across datasets, cross-type spatial interactions and topological descriptors show associations with patient survival, demonstrating that complementary spatial features capture distinct aspects of tumor microenvironment architecture. SASHIMI provides an accessible, reproducible platform for single-cell-level spatial profiling of tumor tissue, enabling interactive visualization and standardized feature extraction without requiring programming expertise.

stat.AP↗

arXiv2025

DIRC-RAG: Accelerating Edge RAG with Robust High-Density and High-Loading-Bandwidth Digital In-ReRAM Computation

Retrieval-Augmented Generation (RAG) enhances large language models (LLMs) by integrating external knowledge retrieval but faces challenges on edge devices due to high storage, energy, and latency demands. Computing-in-Memory (CIM) offers a promising solution by storing document embeddings in CIM macros and enabling in-situ parallel retrievals but is constrained by either low memory density or limited computational accuracy. To address these challenges, we present DIRCRAG, a novel edge RAG acceleration architecture leveraging Digital In-ReRAM Computation (DIRC). DIRC integrates a high-density multi-level ReRAM subarray with an SRAM cell, utilizing SRAM and differential sensing for robust ReRAM readout and digital multiply-accumulate (MAC) operations. By storing all document embeddings within the CIM macro, DIRC achieves ultra-low-power, single-cycle data loading, substantially reducing both energy consumption and latency compared to offchip DRAM. A query-stationary (QS) dataflow is supported for RAG tasks, minimizing on-chip data movement and reducing SRAM buffer requirements. We introduce error optimization for the DIRC ReRAM-SRAM cell by extracting the bit-wise spatial error distribution of the ReRAM subarray and applying targeted bit-wise data remapping. An error detection circuit is also implemented to enhance readout resilience against deviceand circuit-level variations. Simulation results demonstrate that DIRC-RAG under TSMC40nm process achieves an on-chip non-volatile memory density of 5.18Mb/mm2 and a throughput of 131 TOPS. It delivers a 4MB retrieval latency of 5.6{\mu}s/query and an energy consumption of 0.956{\mu}J/query, while maintaining the retrieval precision.

cs.AR↗

arXiv2025

Faster MoE LLM Inference for Extremely Large Models

Sparse Mixture of Experts (MoE) large language models (LLMs) are gradually becoming the mainstream approach for ultra-large-scale models. Existing optimization efforts for MoE models have focused primarily on coarse-grained MoE architectures. With the emergence of DeepSeek Models, fine-grained MoE models are gaining popularity, yet research on them remains limited. Therefore, we want to discuss the efficiency dynamic under different service loads. Additionally, fine-grained models allow deployers to reduce the number of routed experts, both activated counts and total counts, raising the question of how this reduction affects the trade-off between MoE efficiency and performance. Our findings indicate that while deploying MoE models presents greater challenges, it also offers significant optimization opportunities. Reducing the number of activated experts can lead to substantial efficiency improvements in certain scenarios, with only minor performance degradation. Reducing the total number of experts provides limited efficiency gains but results in severe performance degradation. Our method can increase throughput by at least 10\% without any performance degradation. Overall, we conclude that MoE inference optimization remains an area with substantial potential for exploration and improvement.

cs.CL↗

arXiv2025

A Dual-Task Synergy-Driven Generalization Framework for Pancreatic Cancer Segmentation in CT Scans

Pancreatic cancer, characterized by its notable prevalence and mortality rates, demands accurate lesion delineation for effective diagnosis and therapeutic interventions. The generalizability of extant methods is frequently compromised due to the pronounced variability in imaging and the heterogeneous characteristics of pancreatic lesions, which may mimic normal tissues and exhibit significant inter-patient variability. Thus, we propose a generalization framework that synergizes pixel-level classification and regression tasks, to accurately delineate lesions and improve model stability. This framework not only seeks to align segmentation contours with actual lesions but also uses regression to elucidate spatial relationships between diseased and normal tissues, thereby improving tumor localization and morphological characterization. Enhanced by the reciprocal transformation of task outputs, our approach integrates additional regression supervision within the segmentation context, bolstering the model's generalization ability from a dual-task perspective. Besides, dual self-supervised learning in feature spaces and output spaces augments the model's representational capability and stability across different imaging views. Experiments on 594 samples composed of three datasets with significant imaging differences demonstrate that our generalized pancreas segmentation results comparable to mainstream in-domain validation performance (Dice: 84.07%). More importantly, it successfully improves the results of the highly challenging cross-lesion generalized pancreatic cancer segmentation task by 9.51%. Thus, our model constitutes a resilient and efficient foundational technological support for pancreatic disease management and wider medical applications. The codes will be released at https://github.com/SJTUBME-QianLab/Dual-Task-Seg.

eess.IV↗

arXiv2025

BISON: Bi-clustering of spatial omics data with feature selection

The advent of next-generation sequencing-based spatially resolved transcriptomics (SRT) techniques has reshaped genomic studies by enabling high-throughput gene expression profiling while preserving spatial and morphological context. Understanding gene functions and interactions in different spatial domains is crucial, as it can enhance our comprehension of biological mechanisms, such as cancer-immune interactions and cell differentiation in various regions. It is necessary to cluster tissue regions into distinct spatial domains and identify discriminating genes that elucidate the clustering result, referred to as spatial domain-specific discriminating genes (DGs). Existing methods for identifying these genes typically rely on a two-stage approach, which can lead to the phenomenon known as \textit{double-dipping}. To address the challenge, we propose a unified Bayesian latent block model that simultaneously detects a list of DGs contributing to spatial domain identification while clustering these DGs and spatial locations. The efficacy of our proposed method is validated through a series of simulation experiments, and its capability to identify DGs is demonstrated through applications to benchmark SRT datasets.

stat.AP↗

arXiv2024

CAPrompt: Cyclic Prompt Aggregation for Pre-Trained Model Based Class Incremental Learning

Recently, prompt tuning methods for pre-trained models have demonstrated promising performance in Class Incremental Learning (CIL). These methods typically involve learning task-specific prompts and predicting the task ID to select the appropriate prompts for inference. However, inaccurate task ID predictions can cause severe inconsistencies between the prompts used during training and inference, leading to knowledge forgetting and performance degradation. Additionally, existing prompt tuning methods rely solely on the pre-trained model to predict task IDs, without fully leveraging the knowledge embedded in the learned prompt parameters, resulting in inferior prediction performance. To address these issues, we propose a novel Cyclic Prompt Aggregation (CAPrompt) method that eliminates the dependency on task ID prediction by cyclically aggregating the knowledge from different prompts. Specifically, rather than predicting task IDs, we introduce an innovative prompt aggregation strategy during both training and inference to overcome prompt inconsistency by utilizing a weighted sum of different prompts. Thorough theoretical analysis demonstrates that under concave conditions, the aggregated prompt achieves lower error compared to selecting a single task-specific prompt. Consequently, we incorporate a concave constraint and a linear constraint to guide prompt learning, ensuring compliance with the concave condition requirement. Furthermore, to fully exploit the prompts and achieve more accurate prompt weights, we develop a cyclic weight prediction strategy. This strategy begins with equal weights for each task and automatically adjusts them to more appropriate values in a cyclical manner. Experiments on various datasets demonstrate that our proposed CAPrompt outperforms state-of-the-art methods by 2%-3%. Our code is available at https://github.com/zhoujiahuan1991/AAAI2025-CAPrompt.

cs.CV↗

arXiv2024

Generalized Bayesian nonparametric clustering framework for high-dimensional spatial omics data

The advent of next-generation sequencing-based spatially resolved transcriptomics (SRT) techniques has transformed genomic research by enabling high-throughput gene expression profiling while preserving spatial context. Identifying spatial domains within SRT data is a critical task, with numerous computational approaches currently available. However, most existing methods rely on a multi-stage process that involves ad-hoc dimension reduction techniques to manage the high dimensionality of SRT data. These low-dimensional embeddings are then subjected to model-based or distance-based clustering methods. Additionally, many approaches depend on arbitrarily specifying the number of clusters (i.e., spatial domains), which can result in information loss and suboptimal downstream analysis. To address these limitations, we propose a novel Bayesian nonparametric mixture of factor analysis (BNPMFA) model, which incorporates a Markov random field-constrained Gibbs-type prior for partitioning high-dimensional spatial omics data. This new prior effectively integrates the spatial constraints inherent in SRT data while simultaneously inferring cluster membership and determining the optimal number of spatial domains. We have established the theoretical identifiability of cluster membership within this framework. The efficacy of our proposed approach is demonstrated through realistic simulations and applications to two SRT datasets. Our results show that the BNPMFA model not only surpasses state-of-the-art methods in clustering accuracy and estimating the number of clusters but also offers novel insights for identifying cellular regions within tissue samples.

stat.ME↗

arXiv2024

Hypergraph based Understanding for Document Semantic Entity Recognition

Semantic entity recognition is an important task in the field of visually-rich document understanding. It distinguishes the semantic types of text by analyzing the position relationship between text nodes and the relation between text content. The existing document understanding models mainly focus on entity categories while ignoring the extraction of entity boundaries. We build a novel hypergraph attention document semantic entity recognition framework, HGA, which uses hypergraph attention to focus on entity boundaries and entity categories at the same time. It can conduct a more detailed analysis of the document text representation analyzed by the upstream model and achieves a better performance of semantic information. We apply this method on the basis of GraphLayoutLM to construct a new semantic entity recognition model HGALayoutLM. Our experiment results on FUNSD, CORD, XFUND and SROIE show that our method can effectively improve the performance of semantic entity recognition tasks based on the original model. The results of HGALayoutLM on FUNSD and XFUND reach the new state-of-the-art results.

cs.AI↗

arXiv2024

Bayesian modeling of co-occurrence microbial interaction networks

The human body consists of microbiomes associated with the development and prevention of several diseases. These microbial organisms form several complex interactions that are informative to the scientific community for explaining disease progression and prevention. Contrary to the traditional view of the microbiome as a singular, assortative network, we introduce a novel statistical approach using a weighted stochastic infinite block model to analyze the complex community structures within microbial co-occurrence microbial interaction networks. Our model defines connections between microbial taxa using a novel semi-parametric rank-based correlation method on their transformed relative abundances within a fully connected network framework. Employing a Bayesian nonparametric approach, the proposed model effectively clusters taxa into distinct communities while estimating the number of communities. The posterior summary of the taxa community membership is obtained based on the posterior probability matrix, which could naturally solve the label switching problem. Through simulation studies and real-world application to microbiome data from postmenopausal patients with recurrent urinary tract infections, we demonstrate that our method has superior clustering accuracy over alternative approaches. This advancement provides a more nuanced understanding of microbiome organization, with significant implications for disease research.

stat.ME↗

arXiv2023

Bayesian Nonparametric Clustering with Feature Selection for Spatially Resolved Transcriptomics Data

The advent of next-generation sequencing-based spatially resolved transcriptomics (SRT) techniques has reshaped genomic studies by enabling high-throughput gene expression profiling while preserving spatial and morphological context. Nevertheless, there are inherent challenges associated with these new high-dimensional spatial data, such as zero-inflation, over-dispersion, and heterogeneity. These challenges pose obstacles to effective clustering, which is a fundamental problem in SRT data analysis. Current computational approaches often rely on heuristic data preprocessing and arbitrary cluster number prespecification, leading to considerable information loss and consequently, suboptimal downstream analysis. In response to these challenges, we introduce BNPSpace, a novel Bayesian nonparametric spatial clustering framework that directly models SRT count data. BNPSpace facilitates the partitioning of the whole spatial domain, which is characterized by substantial heterogeneity, into homogeneous spatial domains with similar molecular characteristics while identifying a parsimonious set of discriminating genes among different spatial domains. Moreover, BNPSpace incorporates spatial information through a Markov random field prior model, encouraging a smooth and biologically meaningful partition pattern.

stat.AP↗

arXiv2023

A generalized Bayesian stochastic block model for microbiome community detection

Advances in next-generation sequencing technology have enabled the high-throughput profiling of metagenomes and accelerated the microbiome study. Recently, there has been a rise in quantitative studies that aim to decipher the microbiome co-occurrence network and its underlying community structure based on metagenomic sequence data. Uncovering the complex microbiome community structure is essential to understanding the role of the microbiome in disease progression and susceptibility. Taxonomic abundance data generated from metagenomic sequencing technologies are high-dimensional and compositional, suffering from uneven sampling depth, over-dispersion, and zero-inflation. These characteristics often challenge the reliability of the current methods for microbiome community detection. To this end, we propose a Bayesian stochastic block model to study the microbiome co-occurrence network based on the recently developed modified centered-log ratio transformation tailored for microbiome data analysis. Our model allows us to incorporate taxonomic tree information using a Markov random field prior. The model parameters are jointly inferred by using Markov chain Monte Carlo sampling techniques. Our simulation study showed that the proposed approach performs better than competing methods even when taxonomic tree information is non-informative. We applied our approach to a real urinary microbiome dataset from postmenopausal women, the first time the urinary microbiome co-occurrence network structure has been studied. In summary, this statistical methodology provides a new tool for facilitating advanced microbiome studies.

stat.ME↗