arXiv subjects

Chi Zhang

Publications and source records attributed to Chi Zhang.

At least 181 records · Page 10Linked to original sources

arXiv2026

Preserve Support, Not Correspondence: Dynamic Routing for Offline Reinforcement Learning

One-step offline RL actors are attractive because they avoid backpropagating through long iterative samplers and keep inference cheap, but they still have to improve under a critic without drifting away from actions that the dataset can support. In recent one-step extraction pipelines, a strong iterative teacher provides one target action for each latent draw, and the same student output is asked to do both jobs: move toward higher Q and stay near that paired endpoint. If those two directions disagree, the loss resolves them as a compromise on that same sample, even when a nearby better action remains locally supported by the data. We propose DROL, a latent-conditioned one-step actor trained with top-1 dynamic routing. For each state, the actor samples $K$ candidate actions from a bounded latent prior, assigns each dataset action to its nearest candidate, and updates only that winner with Behavior Cloning and critic guidance. Because the routing is recomputed from the current candidate geometry, ownership of a supported region can shift across candidates over the course of learning. This gives a one-step actor room to make local improvements that pointwise extraction struggles to capture, while retaining single-pass inference at test time. On OGBench and D4RL, DROL is competitive with the one-step FQL baseline, improving many OGBench task groups while remaining strong on both AntMaze and Adroit. Project page: https://muzhancun.github.io/preprints/DROL.

cs.LG↗

arXiv2026

Notes on the LVP and CVP in $p$-adic Fields

This paper explores computational methods for solving the Longest Vector Problem (LVP) and Closest Vector Problem (CVP) in $p$-adic fields. Leveraging the non-Archimedean property of $p$-adic norms, we propose a polynomial time algorithm to compute orthogonal bases for $p$-adic lattices when the $p$-adic field is given by a minimal polynomial. The method utilizes the structure of maximal orders and $p$-radicals in extension fields of $\mathbb{Q}_{p}$ to efficiently construct uniformizers and residue field bases, enabling rapid solutions for the LVP and CVP. In addition, we introduce the characterization of norms on vector spaces over $\mathbb{Q}_p$.

math.NT↗

arXiv2026

VTouch++: A Multimodal Dataset with Vision-Based Tactile Enhancement for Bimanual Manipulation

Embodied intelligence has advanced rapidly in recent years; however, bimanual manipulation-especially in contact-rich tasks remains challenging. This is largely due to the lack of datasets with rich physical interaction signals, systematic task organization, and sufficient scale. To address these limitations, we introduce the VTOUCH dataset. It leverages vision based tactile sensing to provide high-fidelity physical interaction signals, adopts a matrix-style task design to enable systematic learning, and employs automated data collection pipelines covering real-world, demand-driven scenarios to ensure scalability. To further validate the effectiveness of the dataset, we conduct extensive quantitative experiments on cross-modal retrieval as well as real-robot evaluation. Finally, we demonstrate real-world performance through generalizable inference across multiple robots, policies, and tasks.

cs.RO↗

arXiv2026

Modeling Multiple Support Strategies within a Single Turn for Emotional Support Conversations

Emotional Support Conversation (ESC) aims to assist individuals experiencing distress by generating empathetic and supportive dialogue. While prior work typically assumes that each supporter turn corresponds to a single strategy, real-world supportive communication often involves multiple strategies within a single utterance. In this paper, we revisit the ESC task by formulating it as multi-strategy utterance generation, where each utterance may contain one or more strategy-response pairs. We propose two generation methods: All-in-One, which predicts all strategy-response pairs in a single decoding step, and One-by-One, which iteratively generates strategy-response pairs until completion. Both methods are further enhanced with cognitive reasoning guided by reinforcement learning to improve strategy selection and response composition. We evaluate our models on the ESConv dataset under both utterance-level and dialogue-level settings. Experimental results show that our methods effectively model multi-strategy utterances and lead to improved supportive quality and dialogue success. To our knowledge, this work provides the first systematic empirical evidence that allowing multiple support strategies within a single utterance is both feasible and beneficial for emotional support conversations. All code and data will be publicly available at https://github.com/aliyun/qwen-dianjin.

cs.CL↗

arXiv2026

ClariCodec: Optimising Neural Speech Codes for 200bps Communication using Reinforcement Learning

In bandwidth-constrained communication such as satellite and underwater channels, speech must often be transmitted at ultra-low bitrates where intelligibility is the primary objective. At such extreme compression levels, codecs trained with acoustic reconstruction losses tend to allocate bits to perceptual detail, leading to substantial degradation in word error rate (WER). This paper proposes ClariCodec, a neural speech codec operating at 200 bit per second (bps) that reformulates quantisation as a stochastic policy, enabling reinforcement learning (RL)-based optimisation of intelligibility. Specifically, the encoder is fine-tuned using WER-driven rewards while the acoustic reconstruction pipeline remains frozen. Even without RL, ClariCodec achieves 3.68% WER on the LibriSpeech test-clean set at 200 bps, already competitive with codecs operating at higher bitrates. Further RL fine-tuning reduces WER to 3.20% on test-clean and 8.93% on test-other, corresponding to a 13% relative reduction while preserving perceptual quality.

cs.SD↗

arXiv2026

Transparent and Controllable Recommendation Filtering via Multimodal Multi-Agent Collaboration

While personalized recommender systems excel at content discovery, they frequently expose users to undesirable or discomforting information, highlighting the critical need for user-centric filtering tools. Current methods leveraging Large Language Models (LLMs) struggle with two major bottlenecks: they lack multimodal awareness to identify visually inappropriate content, and they are highly prone to "over-association" -- incorrectly generalizing a user's specific dislike (e.g., anxiety-inducing marketing) to block benign, educational materials. These unconstrained hallucinations lead to a high volume of false positives, ultimately undermining user agency. To overcome these challenges, we introduce a novel framework that integrates end-to-cloud collaboration, multimodal perception, and multi-agent orchestration. Our system employs a fact-grounded adjudication pipeline to eliminate inferential hallucinations. Furthermore, it constructs a dynamic, two-tier preference graph that allows for explicit, human-in-the-loop modifications (via Delta-adjustments), explicitly preventing the algorithm from catastrophically forgetting fine-grained user intents. Evaluated on an adversarial dataset comprising 473 highly confusing samples, the proposed architecture effectively curbed over-association, decreasing the false positive rate by 74.3% and achieving nearly twice the F1-Score of traditional text-only baselines. Additionally, a 7-day longitudinal field study with 19 participants demonstrated robust intent alignment and enhanced governance efficiency. User feedback confirmed that the framework drastically improves algorithmic transparency, rebuilds user control, and alleviates the fear of missing out (FOMO), paving the way for transparent human-AI co-governance in personalized feeds.

cs.IR↗

arXiv2026

Automated Classification of Plasma Regions at Mars Using Machine Learning

The plasma environment around Mars is highly variable because it is strongly influenced by the solar wind. Accurate identification of plasma regions around Mars is important for the community studying solar wind-Mars interactions, region-specific plasma processes, and atmospheric escape. In this study, we develop a machine-learning-based classifier to automatically identify three key plasma regions--solar wind, magnetosheath, and induced magnetosphere--using only ion omnidirectional energy spectra measured by the MAVEN Solar Wind Ion Analyzer (SWIA). Two neural network architectures are evaluated: a multilayer perceptron (MLP) and a convolutional neural network (CNN) that incorporates short temporal sequences. Our results show that the CNN can reliably distinguish the three plasma regions, whereas the MLP struggles to separate the solar wind and magnetosheath. Therefore, the CNN-based approach provides an efficient and accurate framework for large-scale plasma region identification at Mars and can be readily applied to future planetary missions.

physics.space-ph↗

arXiv2026

AppAgent-Claw: CLI Is All You Need for GUI Automation

The OpenClaw platform provides a practical foundation for automation through its skill-oriented architecture, organizing external capabilities into lightweight, reusable components that can be invoked efficiently through a command-line interface (CLI). However, a significant bottleneck remains: many real-world tasks are confined to graphical user interfaces (GUIs) with no stable API available. While LLM-based GUI agents offer generality, their reliance on repeated live model inference makes them too slow, costly, and inconsistent to serve as efficient OpenClaw skills. In this paper, we present AppAgent-Claw, a demonstration-driven system that converts GUI workflows into reliable, reusable skills without runtime inference. By following a ``record-once, replay-many'' paradigm, the system captures rich contextual metadata to facilitate robust execution. It employs a layered localization strategy to handle visual shifts and a validation-coupled execution model to ensure intended on-screen effects. AppAgent-Claw provides a practical, efficient, and diagnosable solution for integrating GUI-bound tasks into the OpenClaw ecosystem.

cs.HC↗

arXiv2026

CODO: An Automated Compiler for Comprehensive Dataflow Optimization

FPGAs are well-suited for dataflow architectures that process data in a streaming or pipelined manner, thus satisfying the high computational and communication demands of emerging applications. However, manually implementing an efficient dataflow architecture for large-scale applications is still challenging, even for specialists who use high-level synthesis (HLS) to simplify FPGA programming. To address this, we introduce CODO, an automated compiler that generates feasible and efficient dataflow accelerators on FPGAs. CODO features a systematic method for detecting and eliminating both coarse-grained and fine-grained dataflow violations. Building on this, CODO performs both on- and off-chip data movement optimizations to maximize transfer efficiency. To guarantee a higher design quality, CODO performs automatic scheduling to generate high-performance dataflow accelerators, ensuring a balanced performance-resource trade-off. Synthesis results show that CODO delivers $1.45\times$ to $4.52\times$ latency speedups on typical computation kernels and $3.7\times$ to $33.8\times$ speedups on DNN models compared to SOTA frameworks. In on-board evaluations, CODO achieves $7.3\times$ average speedup on CNN models and $2.07\times$ average speedup on the GPT-2 model over SOTA frameworks. The compiler is open-sourced at https://github.com/sjtu-zhao-lab/codo-artifact.

cs.AR↗

arXiv2026

DeepSketcher: Internalizing Visual Manipulation for Multimodal Reasoning

The "thinking with images" paradigm represents a pivotal shift in the reasoning of Vision Language Models (VLMs), moving from text-dominant chain-of-thought to image-interactive reasoning. By invoking visual tools or generating intermediate visual representations, VLMs can iteratively attend to fine-grained regions, enabling deeper image understanding and more faithful multimodal reasoning. As an emerging paradigm, however, it still leaves substantial room for exploration in data construction accuracy, structural design, and broader application scenarios, which offer rich opportunities for advancing multimodal reasoning. To further advance this line of work, we present DeepSketcher, a comprehensive suite comprising both an image-text interleaved dataset and a self-contained model. The dataset contains 31k chain-of-thought (CoT) reasoning trajectories with diverse tool calls and resulting edited images, covering a wide range of data types and manipulation instructions with high annotation accuracy. Building on this resource, we design a model that performs interleaved image-text reasoning and natively generates "visual thoughts" by operating directly in the visual embedding space, rather than invoking external tools and repeatedly re-encoding generated images. This design enables tool-free and more flexible "thinking with images". Extensive experiments on multimodal reasoning benchmarks demonstrate strong performance, validating both the utility of the dataset and the effectiveness of the model design.

cs.CV↗

arXiv2026

Closed-Form Concept Erasure via Double Projections

While modern generative models such as diffusion-based architectures have enabled impressive creative capabilities, they also raise important safety and ethical risks. These concerns have led to growing interest in concept erasure, the process of removing unwanted concepts from model representations. Existing approaches often achieve strong erasure performance but rely on iterative optimization and may inadvertently distort unrelated concepts. In this work, we present a simple yet principled alternative: a linear transformation framework that achieves concept erasure analytically, without any training. Our method adapts a pretrained model through two sequential, closed-form steps: first, computing a proxy projection of the target concept, and second, applying a constrained transformation within the left null space of known concept directions. This design yields a deterministic and geometrically interpretable procedure for safe, efficient, and theory-grounded concept removal. Across a wide range of experiments, including object and style erasure on multiple Stable Diffusion variants and the flow-matching model (FLUX), our approach matches or surpasses the performance of state-of-the-art methods while preserving non-target concepts more faithfully. Requiring only a few seconds to apply, it offers a lightweight and drop-in tool for controlled model editing, advancing the goal of safer and more responsible generative models.

cs.LG↗

arXiv2026

Investigation of coherence of niobium-based resonators enabled by a fast-sealing microwave cavity

Resonators and qubits with a niobium (Nb) base metal layer achieve some of the highest coherence times in superconducting quantum devices. The performance of such devices is often limited by loss associated with two-level systems, which are found primarily at material surfaces and interfaces. The metal-air (MA) interface is a major contributor to device loss. In this work, we develop a fast-sealing microwave cavity that enables devices to be placed under vacuum within five minutes of oxide removal, thereby significantly reducing the MA interface loss compared to common device processing and packaging approaches. Using coplanar stripline resonators, we demonstrate that devices sealed in such a cavity exhibit internal quality factors exceeding one million at single-photon power. After re-exposure to air, the devices show downward resonance frequency shifts and quality factor degradations, quantitatively consistent with a model of Nb oxide regrowth. The fast-sealing microwave cavity provides a practical and consistent method to mitigate MA interface loss and sustain high coherence in Nb devices, and establishes a controlled platform for studying metal oxide regrowth kinetics and dielectric properties, the understanding of which is critical to achieving high coherence in superconducting quantum devices.

quant-ph↗

arXiv2026

BugForge: Constructing and Utilizing DBMS Bug Repository to Enhance DBMS Testing

DBMSs are complex systems prone to bugs that may lead to system failures or compromise data integrity. Establishing unified DBMS bug repositories is crucial for systematically organizing bug-related data, enabling code improvement, and supporting automated testing. In particular, bug reports often contain valuable test inputs and bug-triggering clues that help explore rare execution paths and expose critical buggy behavior, thereby guiding automated DBMS testing. However, the heterogeneity of bug reports, along with their incomplete or inaccurate content, makes it challenging to build unified repositories and convert them into high-quality test cases. In this paper, we propose BugForge, a framework that constructs standardized DBMS bug repositories and leverages them to generate high-quality test cases to enhance DBMS testing. Specifically, BugForge progressively collects bug reports, then employs syntax-aware processing and input-adaptive raw PoC extraction to construct a DBMS bug repository. The repository stores structured bug-related data, including bug metadata and raw PoCs that entail potential bug-triggering semantics. These data are further refined into high-quality test cases through semantic-guided adaptation, thereby enabling enhanced DBMS testing methods, including DBMS fuzzing, regression testing, and cross-DBMS bug discovery. We implemented BugForge for PostgreSQL, MySQL, MariaDB, and MonetDB, totally integrated 37,632 bug reports spanning up to 28 years. Based on the repository, BugForge uncovered 35 previously unknown bugs with 22 confirmed by developers, demonstrating the value of constructing and utilizing bug repositories for DBMS testing.

cs.SE↗

arXiv2026

FlatAttention: Dataflow and Fabric Collectives Co-Optimization for Large Attention-Based Model Inference on Tile-Based Accelerators

Attention accounts for an increasingly dominant fraction of total computation during inference for mixture-of-experts (MoE) models, making efficient acceleration critical. Emerging domain-specific accelerators for large model inference are shifting toward chip-scale and wafer-scale tile-based architectures. Tiles contain large matrix and vector engines and are connected through on-chip interconnects, which support tile-to-tile traffic to reduce the tile-to-main-memory traffic bottleneck. Hence, dataflow management is crucial to achieve high utilization. We propose FlatAttention, a dataflow for modern attention variants on tile-based accelerators. FlatAttention minimizes expensive high-bandwidth memory (HBM) accesses by exploiting collective primitives integrated into the on-chip network fabric, achieving up to 92.3% utilization, 4.1x speedup over FlashAttention-3, and 16x lower HBM traffic. On a 32x32 tile configuration with peak performance comparable to NVIDIA GH200, FlatAttention generalizes across multiple attention variants, achieving an average of 86% utilization for compute-bound attentions and 78% HBM bandwidth utilization for memory-bound ones, resulting in an average 1.9x speedup over attention implementations on GH200. Finally, we evaluate end-to-end DeepSeek-v3 FP8 decoding with FlatAttention on a wafer-scale multi-die system, achieving a 1.9x improvement in system throughput and a 1.4x reduction in per-user token output latency, despite operating with 1.5x lower peak system performance compared to the state-of-the-art solution.

cs.AR↗

arXiv2026

Auto-Slides: An Interactive Multi-Agent System for Creating and Customizing Research Presentations

The rapid progress of large language models (LLMs) has opened new opportunities for education. While learners can interact with academic papers through LLM-powered dialogue, limitations still exist: the lack of structured organization and the heavy reliance on text can impede systematic understanding and engagement with complex concepts. To address these challenges, we propose Auto-Slides, an LLM-driven system that converts research papers into pedagogically structured, multimodal slides (e.g., diagrams and tables). Drawing on cognitive science, it creates a presentation-oriented narrative and allows iterative refinement via an interactive editor to better match learners' knowledge level and goals. Auto-Slides further incorporates verification and knowledge retrieval mechanisms to ensure accuracy and contextual completeness. Through extensive user studies, Auto-Slides demonstrates strong learner acceptance, improved structural support for understanding, and expert-validated gains in narrative quality compared with conventional LLM-based reading. Our contributions lie in designing a multi-agent framework for transforming academic papers into pedagogically optimized slides and introducing interactive customization for personalized learning.

cs.HC↗

arXiv2026

Multipartite controlled-NOT gates using molecules and Rydberg atoms

We propose high-fidelity controlled-NOT (CNOT) gates in a hybrid system of polar molecules and Rydberg atoms based on the unconventional Rydberg pumping mechanism. By combining the rich internal structure of polar molecules with the strong dipole-dipole interactions of Rydberg atoms, we realize both two-to-one and one-to-two gate configurations. Numerical simulations show that the gate performance is robust against spontaneous emission from Rydberg states. The approach naturally extends to larger systems, as demonstrated by four-qubit implementations achieving three-to-one and one-to-three CNOT gates with fidelities exceeding 99\%. These results highlight hybrid molecule-Rydberg atom architectures as a promising platform for scalable quantum information processing.

quant-ph↗

arXiv2026

LogiStory: A Logic-Aware Framework for Multi-Image Story Visualization

Generating coherent and communicative visual sequences, such as image sequences and videos, remains a significant challenge for current multimodal systems. Despite advances in visual quality and the integration of world knowledge, existing models still struggle to maintain logical flow, often resulting in disjointed actions, fragmented narratives, and unclear storylines. We attribute these issues to the lack of attention to visual logic, a critical yet underexplored dimension of visual sequence generation that we define as the perceptual and causal coherence among characters, actions, and scenes over time. To bridge this gap, we propose a logic-aware multi-image story visualization framework, LogiStory. The framework is built around the central innovation of explicitly modeling visual logic in story visualization. To realize this idea, we design a multi-agent system that grounds roles, extracts causal chains, and verifies story-level consistency, transforming narrative coherence from an implicit byproduct of image generation into an explicit modeling objective. This design effectively bridges structured story planning with visual generation, enhancing both narrative clarity and visual quality in story visualization. Furthermore, to evaluate the generation capacity, we construct LogicTale, a benchmark comprising richly annotated stories, emphasizing causal reasoning, and visual logic interpretability. We establish comprehensive automatic and human evaluation protocols designed to measure both visual logic and perceptual quality. Experiments demonstrate that our approach significantly improves the narrative logic of generated visual stories. This work provides a foundational step towards modeling and enforcing visual logic in general image sequence and video generation tasks.

cs.CV↗

arXiv2026

ViStoryBench: Comprehensive Benchmark Suite for Story Visualization

Story visualization aims to generate coherent image sequences that faithfully represent a narrative and match given character references. Despite progress in generative models, existing benchmarks remain narrow in scope, often limited to short prompts, lacking character references, or single-image cases, failing to reflect real-world narrative complexity and obscuring true model performance.We introduce ViStoryBench, a comprehensive benchmark designed to evaluate story visualization models across varied narrative structures, visual styles, and character settings. It features richly annotated multi-shot scripts derived from curated stories spanning literature, film, and folklore. Large language models assist in story summarization and script generation, with all outputs verified by humans for coherence and fidelity. Character references are carefully curated to maintain consistency across different artistic styles. ViStoryBench proposes a suite of multi-dimensional automated metrics to evaluate character consistency, style similarity, prompt alignment, aesthetic quality, and artifacts like copy-paste behavior. These metrics are validated through human studies and used to assess a broad range of open-source and commercial models, enabling systematic analysis and encouraging advances in visual storytelling.

cs.CV↗