arXiv subjects

Kun Qian

Publications and source records attributed to Kun Qian.

At least 37 records · Page 2Linked to original sources

arXiv2026

Human-AI Co-reasoning for Clinical Diagnosis with Evidence-Integrated Language Agent

We present PULSE, a medical reasoning agent that combines a domain-tuned large language model with scientific literature retrieval to support diagnostic decision-making in complex real-world cases. To evaluate its capabilities, we curated a benchmark of 82 authentic endocrinology case reports encompassing a broad spectrum of disease types and incidence levels. In controlled experiments, we compared PULSE's performance against physicians with varying levels of expertise-from residents to senior specialists-and examined how AI assistance influenced human diagnostic reasoning. PULSE attained expert-competitive accuracy, outperforming residents and junior specialists while matching senior specialist performance at both Top@1 and Top@4 thresholds. Unlike physicians, whose accuracy declined with disease rarity, PULSE maintained stable performance across incidence tiers. The agent also exhibited adaptive reasoning, increasing output length with case difficulty in a manner analogous to the longer deliberation observed among expert clinicians. When used collaboratively, PULSE enabled physicians to correct initial errors and broaden diagnostic hypotheses, but also introduced risks of automation bias. The study explores both serial and concurrent collaboration workflows, revealing that PULSE offers robust support across common and rare presentations. These findings underscore both the promise and the limitations of language model-based agents in clinical diagnosis, and offer a framework for evaluating their role in real-world decision-making.

cs.CL↗

arXiv2026

EvoSchema: Towards Text-to-SQL Robustness Against Schema Evolution

Neural text-to-SQL models, which translate natural language questions (NLQs) into SQL queries given a database schema, have achieved remarkable performance. However, database schemas frequently evolve to meet new requirements. Such schema evolution often leads to performance degradation for models trained on static schemas. Existing work either mainly focuses on simply paraphrasing some syntactic or semantic mappings among NLQ, DB and SQL, or lacks a comprehensive and controllable way to investigate the model robustness issue under the schema evolution, which is insufficient when facing the increasingly complex and rich database schema changes in reality, especially in the LLM era. To address the challenges posed by schema evolution, we present EvoSchema, a comprehensive benchmark designed to assess and enhance the robustness of text-to-SQL systems under real-world schema changes. EvoSchema introduces a novel schema evolution taxonomy, encompassing ten perturbation types across columnlevel and table-level modifications, systematically simulating the dynamic nature of database schemas. Through EvoSchema, we conduct an in-depth evaluation spanning different open source and closed-source LLMs, revealing that table-level perturbations have a significantly greater impact on model performance compared to column-level changes. Furthermore, EvoSchema inspires the development of more resilient text-to-SQL systems, in terms of both model training and database design. The models trained on EvoSchema's diverse schema designs can force the model to distinguish the schema difference for the same questions to avoid learning spurious patterns, which demonstrate remarkable robustness compared to those trained on unperturbed data on average. This benchmark offers valuable insights into model behavior and a path forward for designing systems capable of thriving in dynamic, real-world environments.

cs.DB↗

arXiv2026

EROICA: Online Performance Troubleshooting for Large-scale Model Training

Troubleshooting performance problems of large model training (LMT) is immensely challenging, due to unprecedented scales of modern GPU clusters, the complexity of software-hardware interactions, and the data intensity of the training process. Existing troubleshooting approaches designed for traditional distributed systems or datacenter networks fall short and can hardly apply to real-world training systems. In this paper, we present EROICA, the first online troubleshooting system that provides both fine-grained observation based on profiling, and coverage of all machines in GPU clusters, to diagnose performance issues in production, including both hardware and software problems (or the mixture of both). EROICA effectively summarizes runtime behavior patterns of LMT function executions via online profiling, and leverages differential observability to localize the root cause with minimal production impact. EROICA has been deployed as a production service for large-scale GPU clusters of ~100,000 GPUs for 1.5 years. It has diagnosed a variety of difficult performance issues with 97.5% success.

cs.DC↗

arXiv2026

Adaptation of Agentic AI: A Survey of Post-Training, Memory, and Skills

Large language model (LLM) agents are moving beyond prompting alone. ChatGPT marked the rise of general-purpose LLM assistants, DeepSeek showed that on-policy reinforcement learning with verifiable rewards can improve reasoning and tool use, and OpenClaw highlights a newer direction in which agents accumulate persistent memory and reusable skills. Yet the research landscape remains fragmented across post-training, retrieval, memory, and skill systems. This survey studies these developments under a single notion of \emph{adaptation}: improving an agent, its tools, or their interaction after pretraining. We organize the field with a four-paradigm framework spanning agent adaptation and tool adaptation. On the agent side, A1 (tool-execution-signaled) and A2 (agent-output-signaled) improve the agent itself through supervised fine-tuning, preference optimization, and reinforcement learning with verifiable rewards. On the tool side, T1 (agent-agnostic) provides reusable pre-trained modules any agent can call, while T2 (agent-supervised) uses the agent's outputs to train memory systems, skill libraries, or lightweight subagents. Using this framework, we review post-training methods, adaptive memory architectures, and agent skills; compare their trade-offs in cost, flexibility, and generalization; and summarize evaluation practices across deep research, software development, computer use, and drug discovery. We conclude by outlining open problems in agent-tool co-adaptation, continual learning, safety, and efficient deployment.

cs.AI↗

arXiv2026

Shop-R1: Rewarding LLMs to Simulate Human Behavior in Online Shopping via Reinforcement Learning

Large Language Models (LLMs) have recently demonstrated strong potential in generating 'believable human-like' behavior in web environments. Prior work has explored augmenting training data with LLM-synthesized rationales and applying supervised fine-tuning (SFT) to enhance reasoning ability, which in turn can improve downstream action prediction. However, the performance of such approaches remains inherently bounded by the reasoning capabilities of the model used to generate the rationales. In this paper, we introduce Shop-R1, a novel reinforcement learning (RL) framework aimed at enhancing the reasoning ability of LLMs for simulation of real human behavior in online shopping environments. Specifically, Shop-R1 decomposes the human behavior simulation task into two stages: rationale generation and action prediction, each guided by distinct reward signals. For rationale generation, we leverage internal model signals (e.g., logit distributions) to guide the reasoning process in a self-supervised manner. For action prediction, we propose a hierarchical reward structure with difficulty-aware scaling to prevent reward hacking and enable fine-grained reward assignment. This design evaluates both high-level action types and the correctness of fine-grained sub-action details (attributes and values), rewarding outputs proportionally to their difficulty. Experimental results show that our method achieves a relative improvement of over 65% compared to the baseline. The project page is available at https://damon-demon.github.io/shop-r1.html.

cs.CL↗

arXiv2026

Rec-R1: Bridging Generative Large Language Models and User-Centric Recommendation Systems via Reinforcement Learning

We propose Rec-R1, a general reinforcement learning framework that bridges large language models (LLMs) with recommendation systems through closed-loop optimization. Unlike prompting and supervised fine-tuning (SFT), Rec-R1 directly optimizes LLM generation using feedback from a fixed black-box recommendation model, without relying on synthetic SFT data from proprietary models such as GPT-4o. This avoids the substantial cost and effort required for data distillation. To verify the effectiveness of Rec-R1, we evaluate it on two representative tasks: product search and sequential recommendation. Experimental results demonstrate that Rec-R1 not only consistently outperforms prompting- and SFT-based methods, but also achieves significant gains over strong discriminative baselines, even when used with simple retrievers such as BM25. Moreover, Rec-R1 preserves the general-purpose capabilities of the LLM, unlike SFT, which often impairs instruction-following and reasoning. These findings suggest Rec-R1 as a promising foundation for continual task-specific adaptation without catastrophic forgetting.

cs.IR↗

arXiv2026

SFT Doesn't Always Hurt General Capabilities: Revisiting Domain-Specific Fine-Tuning in LLMs

Supervised Fine-Tuning (SFT) on domain-specific datasets is a common approach to adapt Large Language Models (LLMs) to specialized tasks but is often believed to degrade their general capabilities. In this work, we revisit this trade-off and present both empirical and theoretical insights. First, we show that SFT does not always hurt: using a smaller learning rate can substantially mitigate general performance degradation while preserving comparable target-domain performance. We then provide a theoretical analysis that explains these phenomena and further motivates a new method, Token-Adaptive Loss Reweighting (TALR). Building on this, and recognizing that smaller learning rates alone do not fully eliminate general-performance degradation in all cases, we evaluate a range of strategies for reducing general capability loss, including L2 regularization, LoRA, model averaging, FLOW, and our proposed TALR. Experimental results demonstrate that while no method completely eliminates the trade-off, TALR consistently outperforms these baselines in balancing domain-specific gains and general capabilities. Finally, we distill our findings into practical guidelines for adapting LLMs to new domains: (i) using a small learning rate to achieve a favorable trade-off, and (ii) when a stronger balance is further desired, adopt TALR as an effective strategy.

cs.CL↗

arXiv2025

Spin-flop driven interfacial tunneling magnetoresistance in an antiferromagnetic tunnel junction

The utilization of two-dimensional (2D) materials in magnetic tunnel junctions (MTJs) has shown excellent performance and rich physics. As for 2D antiferromagnets, the magnetic moments in different layers respond asynchronously and can be configured at various states under different magnetic fields, showing the possibility of efficient magnetic and electrical tunability. In this report, A-type antiferromagnetic (AFM) material (Fe0.5Co0.5)5GeTe2 (FCGT) works as electrodes to realize full van der Waals magnetic tunnel junctions. Owing to the interfacial effect, the even-layer FCGT, although with zero net magnetization, exhibits spin selectivity in MTJ architecture contributing to a tunneling magnetoresistance (TMR) reaching about 25% at a low operating current 1 nA at 100 K and persists near room temperature. Due to the surface spin-flop (SSF) effect in antiferromagnetic FCGT, the alternation flexibility between the volatile and nonvolatile memory behavior is achieved. The interfacial TMR can be tuned efficiently in amplitude and even sign under different bias currents and temperatures. These findings show precise magnetoelectric manipulation in MTJs based on 2D antiferromagnets and highlight the promise of 2D antiferromagnets for spintronic devices.

cond-mat.mes-hall↗

arXiv2025

Text-to-SQL Domain Adaptation via Human-LLM Collaborative Data Annotation

Text-to-SQL models, which parse natural language (NL) questions to executable SQL queries, are increasingly adopted in real-world applications. However, deploying such models in the real world often requires adapting them to the highly specialized database schemas used in specific applications. We find that existing text-to-SQL models experience significant performance drops when applied to new schemas, primarily due to the lack of domain-specific data for fine-tuning. This data scarcity also limits the ability to effectively evaluate model performance in new domains. Continuously obtaining high-quality text-to-SQL data for evolving schemas is prohibitively expensive in real-world scenarios. To bridge this gap, we propose SQLsynth, a human-in-the-loop text-to-SQL data annotation system. SQLsynth streamlines the creation of high-quality text-to-SQL datasets through human-LLM collaboration in a structured workflow. A within-subjects user study comparing SQLsynth with manual annotation and ChatGPT shows that SQLsynth significantly accelerates text-to-SQL data annotation, reduces cognitive load, and produces datasets that are more accurate, natural, and diverse. Our code is available at https://github.com/magic-YuanTian/SQLsynth.

cs.HC↗

arXiv2025

Adobe Summit Concierge Evaluation with Human in the Loop

Generative AI assistants offer significant potential to enhance productivity, streamline information access, and improve user experience in enterprise contexts. In this work, we present Summit Concierge, a domain-specific AI assistant developed for Adobe Summit. The assistant handles a wide range of event-related queries and operates under real-world constraints such as data sparsity, quality assurance, and rapid deployment. To address these challenges, we adopt a human-in-the-loop development workflow that combines prompt engineering, retrieval grounding, and lightweight human validation. We describe the system architecture, development process, and real-world deployment outcomes. Our experience shows that agile, feedback-driven development enables scalable and reliable AI assistants, even in cold-start scenarios.

cs.AI↗

arXiv2025

Can any model be fabricated? Inverse operation based planning for hybrid additive-subtractive manufacturing

This paper presents a method for computing interleaved additive and subtractive manufacturing operations to fabricate models of arbitrary shapes. We solve the manufacturing planning problem by searching a sequence of inverse operations that progressively transform a target model into a null shape. Each inverse operation corresponds to either an additive or a subtractive step, ensuring both manufacturability and structural stability of intermediate shapes throughout the process. We theoretically prove that any model can be fabricated exactly using a sequence generated by our approach. To demonstrate the effectiveness of this method, we adopt a voxel-based implementation and develop a scalable algorithm that works on models represented by a large number of voxels. Our approach has been tested across a range of digital models and further validated through physical fabrication on a hybrid manufacturing system with automatic tool switching.

cs.GR↗

arXiv2025

DocRefine: An Intelligent Framework for Scientific Document Understanding and Content Optimization based on Multimodal Large Model Agents

The exponential growth of scientific literature in PDF format necessitates advanced tools for efficient and accurate document understanding, summarization, and content optimization. Traditional methods fall short in handling complex layouts and multimodal content, while direct application of Large Language Models (LLMs) and Vision-Language Large Models (LVLMs) lacks precision and control for intricate editing tasks. This paper introduces DocRefine, an innovative framework designed for intelligent understanding, content refinement, and automated summarization of scientific PDF documents, driven by natural language instructions. DocRefine leverages the power of advanced LVLMs (e.g., GPT-4o) by orchestrating a sophisticated multi-agent system comprising six specialized and collaborative agents: Layout & Structure Analysis, Multimodal Content Understanding, Instruction Decomposition, Content Refinement, Summarization & Generation, and Fidelity & Consistency Verification. This closed-loop feedback architecture ensures high semantic accuracy and visual fidelity. Evaluated on the comprehensive DocEditBench dataset, DocRefine consistently outperforms state-of-the-art baselines across various tasks, achieving overall scores of 86.7% for Semantic Consistency Score (SCS), 93.9% for Layout Fidelity Index (LFI), and 85.0% for Instruction Adherence Rate (IAR). These results demonstrate DocRefine's superior capability in handling complex multimodal document editing, preserving semantic integrity, and maintaining visual consistency, marking a significant advancement in automated scientific document processing.

cs.CV↗

arXiv2025

Sensor-Space Based Robust Kinematic Control of Redundant Soft Manipulator by Learning

The intrinsic compliance and high degree of freedom (DoF) of redundant soft manipulators facilitate safe interaction and flexible task execution. However, effective kinematic control remains highly challenging, as it must handle deformations caused by unknown external loads and avoid actuator saturation due to improper null-space regulation - particularly in confined environments. In this paper, we propose a Sensor-Space Imitation Learning Kinematic Control (SS-ILKC) framework to enable robust kinematic control under actuator saturation and restrictive environmental constraints. We employ a dual-learning strategy: a multi-goal sensor-space control framework based on reinforcement learning principle is trained in simulation to develop robust control policies for open spaces, while a generative adversarial imitation learning approach enables effective policy learning from sparse expert demonstrations for confined spaces. To enable zero-shot real-world deployment, a pre-processed sim-to-real transfer mechanism is proposed to mitigate the simulation-to-reality gap and accurately characterize actuator saturation limits. Experimental results demonstrate that our method can effectively control a pneumatically actuated soft manipulator, achieving precise path-following and object manipulation in confined environments under unknown loading conditions.

cs.RO↗

arXiv2025

Cryogenic in-memory computing using magnetic topological insulators

Machine learning algorithms have been proven effective for essential quantum computation tasks such as quantum error correction and quantum control. Efficient hardware implementation of these algorithms at cryogenic temperatures is essential. Here, we utilize magnetic topological insulators as memristors (termed magnetic topological memristors) and introduce a cryogenic in-memory computing scheme based on the coexistence of the chiral edge state and the topological surface state. The memristive switching and reading of the giant anomalous Hall effect exhibit high energy efficiency, high stability, and low stochasticity. We achieve high accuracy in a proof-of-concept classification task using four magnetic topological memristors. Furthermore, our algorithm-level and circuit-level simulations of large-scale neural networks demonstrate software-level accuracy and lower energy consumption for image recognition and quantum state preparation compared with existing magnetic memristor and CMOS technologies. Our results not only showcase a new application of chiral edge states but also may inspire further topological quantum physics-based novel computing schemes.

cond-mat.mes-hall↗

arXiv2025

MADUV: The 1st INTERSPEECH Mice Autism Detection via Ultrasound Vocalization Challenge

The Mice Autism Detection via Ultrasound Vocalization (MADUV) Challenge introduces the first INTERSPEECH challenge focused on detecting autism spectrum disorder (ASD) in mice through their vocalizations. Participants are tasked with developing models to automatically classify mice as either wild-type or ASD models based on recordings with a high sampling rate. Our baseline system employs a simple CNN-based classification using three different spectrogram features. Results demonstrate the feasibility of automated ASD detection, with the considered audible-range features achieving the best performance (UAR of 0.600 for segment-level and 0.625 for subject-level classification). This challenge bridges speech technology and biomedical research, offering opportunities to advance our understanding of ASD models through machine learning approaches. The findings suggest promising directions for vocalization analysis and highlight the potential value of audible and ultrasound vocalizations in ASD detection.

cs.SD↗

arXiv2025

Breaking Resource Barriers in Speech Emotion Recognition via Data Distillation

Speech emotion recognition (SER) plays a crucial role in human-computer interaction. The emergence of edge devices in the Internet of Things (IoT) presents challenges in constructing intricate deep learning models due to constraints in memory and computational resources. Moreover, emotional speech data often contains private information, raising concerns about privacy leakage during the deployment of SER models. To address these challenges, we propose a data distillation framework to facilitate efficient development of SER models in IoT applications using a synthesised, smaller, and distilled dataset. Our experiments demonstrate that the distilled dataset can be effectively utilised to train SER models with fixed initialisation, achieving performances comparable to those developed using the original full emotional speech dataset.

cs.SD↗

arXiv2025

DECIDER: A Dual-System Rule-Controllable Decoding Framework for Language Generation

Constrained decoding approaches aim to control the meaning or style of text generated by the pre-trained large language models (LLMs or also PLMs) for various tasks at inference time. However, these methods often guide plausible continuations by greedily and explicitly selecting targets. Though fulfilling the task requirements, these methods may overlook certain general and natural logics that humans would implicitly follow towards such targets. Inspired by cognitive dual-process theory, in this work, we propose a novel decoding framework DECIDER where the base LLMs are equipped with a First-Order Logic (FOL) reasoner to express and evaluate the rules, along with a decision function that merges the outputs of both systems to guide the generation. Unlike previous constrained decodings, DECIDER transforms the encouragement of target-specific words into all words that satisfy several high-level rules, enabling us to programmatically integrate our logic into LLMs. Experiments on CommonGen and PersonaChat demonstrate that DECIDER effectively follows given FOL rules to guide LLMs in a more human-like and logic-controlled manner.

cs.CL↗

arXiv2025

DeltaDPD: Exploiting Dynamic Temporal Sparsity in Recurrent Neural Networks for Energy-Efficient Wideband Digital Predistortion

Digital Predistortion (DPD) is a popular technique to enhance signal quality in wideband RF power amplifiers (PAs). With increasing bandwidth and data rates, DPD faces significant energy consumption challenges during deployment, contrasting with its efficiency goals. State-of-the-art DPD models rely on recurrent neural networks (RNN), whose computational complexity hinders system efficiency. This paper introduces DeltaDPD, exploring the dynamic temporal sparsity of input signals and neuronal hidden states in RNNs for energy-efficient DPD, reducing arithmetic operations and memory accesses while preserving satisfactory linearization performance. Applying a TM3.1a 200MHz-BW 256-QAM OFDM signal to a 3.5 GHz GaN Doherty RF PA, DeltaDPD achieves -50.03 dBc in Adjacent Channel Power Ratio (ACPR), -37.22 dB in Normalized Mean Square Error (NMSE) and -38.52 dBc in Error Vector Magnitude (EVM) with 52% temporal sparsity, leading to a 1.8X reduction in estimated inference power. The DeltaDPD code will be released after formal publication at https://www.opendpd.com.

eess.SP↗