arXiv subjects

Jing Xie

Publications and source records attributed to Jing Xie.

At least 19 recordsLinked to original sources

arXiv2026

Recursive Agentic Reasoning

Test-time reasoning methods such as iterative refinement, decomposition, and repeated sampling are often evaluated in isolation, making their gains difficult to compare across models, benchmarks, and evaluation pipelines. We introduce a unified view of these methods as recursion operators over an agent's reasoning trace: GROW, which deepens a single reasoning path; PRUNE, which decomposes and recomposes the problem; and BRANCH, which samples alternative reasoning paths and selects among them. We evaluate all three operators against a single-pass chain-of-thought baseline under a shared harness with identical prompts, token budgets, and grading code. Across five benchmarks and three frontier models, comprising 14 model-benchmark settings, 49,327 graded items, and 151,876 model calls, BRANCH improves accuracy in all 14 settings by an average of 5.98 percentage points and is the best-performing operator in 12. In contrast, GROW yields a mean gain of 2.18 points and degrades performance in two settings, while PRUNE improves accuracy by 0.94 points on average. Analysis shows that BRANCH's advantage arises not only from exploring multiple reasoning paths, but also from recovering from truncation: its gains strongly correlate with the baseline rate of empty, budget-exhausted outputs (r = 0.72). These results weaken the hypothesis that different problems require routing among test-time reasoning operators; at this level of abstraction, repeated branching is consistently dominant. Finally, we show that unpaired evaluation and treating scoring-pipeline failures as model errors can materially change, and even reverse, comparative conclusions, motivating paired scoring as a standard protocol for test-time-compute evaluation.

cs.AI↗

arXiv2026

COMFYCLAW: Self-Evolving Skill Harnesses for Image Generation Workflows

Agents are increasingly used to construct workflows and assist humans in completing recurring tasks more efficiently. As these workflows become repeated and domain-specific, agent memory and reusable skills become increasingly important: agents should be able to recall workflow patterns, execution constraints, and user preferences from previous runs. We study this problem in workflow-based image generation and introduce COMFYCLAW, an agentic skill evolution harness for controlling ComfyUI workflows. COMFYCLAW formulates workflow construction as typed graph editing, exposes tools organized by construction stage, automatically reverts invalid edits, and uses a region-level vision-language model (VLM) verifier to translate visual failures into actionable repair suggestions. The framework further evolves a progressively disclosed skill library, where trajectories, execution errors, and verifier feedback from previous runs are distilled into reusable Agent Skills. Across four benchmark splits, three agent models, and two image backbones, COMFYCLAW achieves the best average image-generation evaluation score across all six agent configurations, outperforming a verifier-only baseline without skill evolution. Human annotations further show that annotators prefer COMFYCLAW over variants without skill evolution. Our results suggest that skill evolution is an effective mechanism for improving agent reliability and performance in recurring visual workflow construction.

cs.AI↗

arXiv2026

Interfacial Electric Fields in Water Nanodroplets are Weakly Dependent on Curvature and pH

The origin of enhanced reactivity in aqueous microdroplets remains debated, with interfacial electric fields (IEFs) often invoked as catalytic drivers. Here, we provide a quantum-mechanical, spatially resolved characterization of the electric field at air-water interfaces by combining deep-learning molecular dynamics with \emph{ab initio} re-sampling. Across planar interfaces and nanodroplets of varying curvature and charge state, we find an outward-oriented field of $\sim 1.0$--$1.2$ V/{\AA} along the intrinsic surface normal. Crucially, its magnitude scales linearly with the average number of hydrogen bonds per interfacial molecule, directly tying the field to the local hydrogen-bond network. Despite its large magnitude and contrary to common expectations, we find that curvature and pH exert only a minor influence on the IEF, becoming negligible at experimentally relevant droplet sizes and pH. Consequently, the reactivity differences observed in $\mu$m-sized droplets cannot be ascribed to variations in the IEF, which changes by a factor of only $\sim10^{-5}$ between $3$ and $40\mu$m-sized droplets. Moreover, the IEF is localized inside the interfacial region and rapidly vanishes within a few {\AA}. This strong spatial confinement renders the IEF strongly tied to the local electronic structure, identifying it as a local property of the air-water boundary rather than an independent physical driver of ``on-water'' catalysis.

physics.chem-ph↗

arXiv2026

PPGL-Swarm: Integrated Multimodal Risk Stratification and Hereditary Syndrome Detection in Pheochromocytoma and Paraganglioma

Pheochromocytomas and paragangliomas (PPGLs) are rare neuroendocrine tumors, of which 15-25% develop metastatic disease with 5-year survival rates reported as low as 34%. PPGL may indicate hereditary syndromes requiring stricter, syndrome-specific treatment and surveillance, but clinicians often fail to recognize these associations in routine care. Clinical practice uses GAPP score for PPGL grading, but several limitations remain for PPGL diagnosis: (1) GAPP scoring demands a high workload for clinician because it requires the manual evaluation of six independent components; (2) key components such as cellularity and Ki-67 are often evaluated with subjective criteria; (3) several clinically relevant metastatic risk factors are not captured by GAPP, such as SDHB mutations, which have been associated with reported metastatic rates of 35-75%. Agent-driven diagnostic systems appear promising, but most lack traceable reasoning for decision-making and do not incorporate domain-specific knowledge such as PPGL genotype information. To address these limitations, we present PPGL-Swarm, an agentic PPGL diagnostic system that generates a comprehensive report, including automated GAPP scoring (with quantified cellularity and Ki-67), genotype risk alerts, and multimodal report with integrated evidence. The system provides an auditable reasoning trail by decomposing diagnosis into micro-tasks, each assigned to a specialized agent. The gene and table agents use knowledge enhancement to better interpret genotype and laboratory findings, and during training we use reinforcement learning to refine tool selection and task assignment.

cs.CV↗

arXiv2026

MM-Zero: Self-Evolving Multi-Model Vision Language Models From Zero Data

Self-evolving has emerged as a key paradigm for improving foundational models such as Large Language Models (LLMs) and Vision Language Models (VLMs) with minimal human intervention. While recent approaches have demonstrated that LLM agents can self-evolve from scratch with little to no data, VLMs introduce an additional visual modality that typically requires at least some seed data, such as images, to bootstrap the self-evolution process. In this work, we present Multi-model Multimodal Zero (MM-Zero), the first RL-based framework to achieve zero-data self-evolution for VLM reasoning. Moving beyond prior dual-role (Proposer and Solver) setups, MM-Zero introduces a multi-role self-evolving training framework comprising three specialized roles: a Proposer that generates abstract visual concepts and formulates questions; a Coder that translates these concepts into executable code (e.g., Python, SVG) to render visual images; and a Solver that performs multimodal reasoning over the generated visual content. All three roles are initialized from the same base model and trained using Group Relative Policy Optimization (GRPO), with carefully designed reward mechanisms that integrate execution feedback, visual verification, and difficulty balancing. Our experiments show that MM-Zero improves VLM reasoning performance across a wide range of multimodal benchmarks. MM-Zero establishes a scalable path toward self-evolving multi-model systems for multimodal models, extending the frontier of self-improvement beyond the conventional two-model paradigm.

cs.CV↗

arXiv2025

Echoes of the Land: An Interactive Installation Based on Physical Model of Earthquake

Echoes of the Land is an interactive installation that transforms seismic dynamics into a multisensory experience through a scientifically grounded spring-block model. Simulating earthquake recurrence and self-organized criticality, the work generates real-time sound and light via motion capture and concatenative granular synthesis. Each block acts as an agent, producing emergent audiovisual cascades that visualize the physics of rupture and threshold behavior. This work exemplifies the amalgamation of scientific knowledge and artistic practice, opening new avenues for novel forms of musical instrument and narrative medium, while inviting further investigation into the intersection of emergent complexity, aesthetics and interactivity.

cs.HC↗

arXiv2024

Enhancing Incremental Summarization with Structured Representations

Large language models (LLMs) often struggle with processing extensive input contexts, which can lead to redundant, inaccurate, or incoherent summaries. Recent methods have used unstructured memory to incrementally process these contexts, but they still suffer from information overload due to the volume of unstructured data handled. In our study, we introduce structured knowledge representations ($GU_{json}$), which significantly improve summarization performance by 40% and 14% across two public datasets. Most notably, we propose the Chain-of-Key strategy ($CoK_{json}$) that dynamically updates or augments these representations with new information, rather than recreating the structured memory for each new source. This method further enhances performance by 7% and 4% on the datasets.

cs.CL↗

arXiv2024

From ChatGPT, DALL-E 3 to Sora: How has Generative AI Changed Digital Humanities Research and Services?

Generative large-scale language models create the fifth paradigm of scientific research, organically combine data science and computational intelligence, transform the research paradigm of natural language processing and multimodal information processing, promote the new trend of AI-enabled social science research, and provide new ideas for digital humanities research and application. This article profoundly explores the application of large-scale language models in digital humanities research, revealing their significant potential in ancient book protection, intelligent processing, and academic innovation. The article first outlines the importance of ancient book resources and the necessity of digital preservation, followed by a detailed introduction to developing large-scale language models, such as ChatGPT, and their applications in document management, content understanding, and cross-cultural research. Through specific cases, the article demonstrates how AI can assist in the organization, classification, and content generation of ancient books. Then, it explores the prospects of AI applications in artistic innovation and cultural heritage preservation. Finally, the article explores the challenges and opportunities in the interaction of technology, information, and society in the digital humanities triggered by AI technologies.

cs.DL↗

arXiv2024

STRUM-LLM: Attributed and Structured Contrastive Summarization

Users often struggle with decision-making between two options (A vs B), as it usually requires time-consuming research across multiple web pages. We propose STRUM-LLM that addresses this challenge by generating attributed, structured, and helpful contrastive summaries that highlight key differences between the two options. STRUM-LLM identifies helpful contrast: the specific attributes along which the two options differ significantly and which are most likely to influence the user's decision. Our technique is domain-agnostic, and does not require any human-labeled data or fixed attribute list as supervision. STRUM-LLM attributes all extractions back to the input sources along with textual evidence, and it does not have a limit on the length of input sources that it can process. STRUM-LLM Distilled has 100x more throughput than the models with comparable performance while being 10x smaller. In this paper, we provide extensive evaluations for our method and lay out future directions for our currently deployed system.

cs.CL↗

arXiv2024

A Learning-based Model Predictive Control Scheme with Application to Temperature Control Units

Temperature control is a complex task due to its often unknown dynamics and disturbances. This paper explores the use of Neural Nonlinear AutoRegressive eXogenous (NNARX) models for nonlinear system identification and model predictive control of a temperature control unit. First, the NNARX model is identified from input-output data collected from the real plant, and a state-space representation with known measurable states consisting of past input and output variables is formulated. Second, a tailored model predictive controller is designed based on the trained NNARX network. The proposed control architecture is experimentally tested on the temperature control units manufactured by Tool-Temp AG. The results achieved are compared with those obtained using a PI controller and a linear MPC. The findings illustrate that the proposed scheme achieves satisfactory tracking performance while incurring the lowest energy cost among the compared controllers.

eess.SY↗

arXiv2024

Internal Model Control design for systems learned by Control Affine Neural Nonlinear Autoregressive Exogenous Models

This paper explores the use of Control Affine Neural Nonlinear AutoRegressive eXogenous (CA-NNARX) models for nonlinear system identification and model-based control design. The idea behind this architecture is to match the known control-affine structure of the system to achieve improved performance. Coherently with recent literature of neural networks for data-driven control, we first analyze the stability properties of CA-NNARX models, devising sufficient conditions for their incremental Input-to-State Stability ($\delta$ISS) that can be enforced at the model training stage. The model's stability property is then leveraged to design a stable Internal Model Control (IMC) architecture. The proposed control scheme is tested on a real Quadruple Tank benchmark system to address the output reference tracking problem. The results achieved show that (i) the modeling accuracy of CA-NNARX is superior to the one of a standard NNARX model for given weight size and training epochs, (ii) the proposed IMC law provides performance comparable to the ones of a standard Model Predictive Controller (MPC) at a significantly lower computational burden, and (iii) the $\delta$ISS of the model is beneficial to the closed-loop performance.

eess.SY↗

arXiv2023

Robust offset-free nonlinear model predictive control for systems learned by neural nonlinear autoregressive exogenous models

This paper presents a robust Model Predictive Control (MPC) scheme that provides offset-free setpoint tracking for systems described by Neural Nonlinear AutoRegressive eXogenous (NNARX) models. The NNARX model learns the dynamics of the plant from input-output data, and during the training the Incremental Input-to-State Stability ($δ$ISS) property is forced to guarantee stability. The trained NNARX model is then augmented with an explicit integral action on the output tracking error, which allows the control scheme to enjoy offset-free tracking ability. A tube-based MPC is finally designed, leveraging the unique structure of the model, to ensure robust stability and robust asymptotic zero error regulation for constant reference signals in the presence of model-plant mismatch or unknown disturbances. Numerical simulations on a water heating system show the effectiveness of the proposed control algorithm.

eess.SY↗

arXiv2023

An Offset-Free Nonlinear MPC scheme for systems learned by Neural NARX models

This paper deals with the design of nonlinear MPC controllers that provide offset-free setpoint tracking for models described by Neural Nonlinear AutoRegressive eXogenous (NNARX) networks. The NNARX model is identified from input-output data collected from the plant, and can be given a state-space representation with known measurable states made by past input and output variables, so that a state observer is not required. In the training phase, the Incremental Input-to-State Stability (δISS) property can be forced when consistent with the behavior of the plant. The δISS property is then leveraged to augment the model with an explicit integral action on the output tracking error, which allows to achieve offset-free tracking capabilities to the designed control scheme. The proposed control architecture is numerically tested on a water heating system and the achieved results are compared to those scored by another popular offset-free MPC method, showing that the proposed scheme attains remarkable performances even in presence of disturbances acting on the plant.

eess.SY↗

arXiv2022

An Augmentation Strategy for Visually Rich Documents

Many business workflows require extracting important fields from form-like documents (e.g. bank statements, bills of lading, purchase orders, etc.). Recent techniques for automating this task work well only when trained with large datasets. In this work we propose a novel data augmentation technique to improve performance when training data is scarce, e.g. 10-250 documents. Our technique, which we call FieldSwap, works by swapping out the key phrases of a source field with the key phrases of a target field to generate new synthetic examples of the target field for use in training. We demonstrate that this approach can yield 1-7 F1 point improvements in extraction performance.

cs.CL↗

arXiv2022

Radically Lower Data-Labeling Costs for Visually Rich Document Extraction Models

A key bottleneck in building automatic extraction models for visually rich documents like invoices is the cost of acquiring the several thousand high-quality labeled documents that are needed to train a model with acceptable accuracy. We propose Selective Labeling to simplify the labeling task to provide "yes/no" labels for candidate extractions predicted by a model trained on partially labeled documents. We combine this with a custom active learning strategy to find the predictions that the model is most uncertain about. We show through experiments on document types drawn from 3 different domains that selective labeling can reduce the cost of acquiring labeled data by $10\times$ with a negligible loss in accuracy.

cs.CL↗

arXiv2022

Towards lifelong learning of Recurrent Neural Networks for control design

This paper proposes a method for lifelong learning of Recurrent Neural Networks, such as NNARX, ESN, LSTM, and GRU, to be used as plant models in control system synthesis. The problem is significant because in many practical applications it is required to adapt the model when new information is available and/or the system undergoes changes, without the need to store an increasing amount of data as time proceeds. Indeed, in this context, many problems arise, such as the well known Catastrophic Forgetting and Capacity Saturation ones. We propose an adaptation algorithm inspired by Moving Horizon Estimators, deriving conditions for its convergence. The described method is applied to a simulated chemical plant, already adopted as a challenging benchmark in the existing literature. The main results achieved are discussed.

eess.SY↗

arXiv2022

On Recurrent Neural Networks for learning-based control: recent results and ideas for future developments

This paper aims to discuss and analyze the potentialities of Recurrent Neural Networks (RNN) in control design applications. The main families of RNN are considered, namely Neural Nonlinear AutoRegressive eXogenous, (NNARX), Echo State Networks (ESN), Long Short Term Memory (LSTM), and Gated Recurrent Units (GRU). The goal is twofold. Firstly, to survey recent results concerning the training of RNN that enjoy Input-to-State Stability (ISS) and Incremental Input-to-State Stability ($δ$ISS) guarantees. Secondly, to discuss the issues that still hinder the widespread use of RNN for control, namely their robustness, verifiability, and interpretability. The former properties are related to the so-called generalization capabilities of the networks, i.e. their consistency with the underlying real plants, even in presence of unseen or perturbed input trajectories. The latter is instead related to the possibility of providing a clear formal connection between the RNN model and the plant. In this context, we illustrate how ISS and $δ$ISS represent a significant step towards the robustness and verifiability of the RNN models, while the requirement of interpretability paves the way to the use of physics-based networks. The design of model predictive controllers with RNN as plant's model is also briefly discussed. Lastly, some of the main topics of the paper are illustrated on a simulated chemical system.

eess.SY↗

arXiv2022

Data-Efficient Information Extraction from Form-Like Documents

Automating information extraction from form-like documents at scale is a pressing need due to its potential impact on automating business workflows across many industries like financial services, insurance, and healthcare. The key challenge is that form-like documents in these business workflows can be laid out in virtually infinitely many ways; hence, a good solution to this problem should generalize to documents with unseen layouts and languages. A solution to this problem requires a holistic understanding of both the textual segments and the visual cues within a document, which is non-trivial. While the natural language processing and computer vision communities are starting to tackle this problem, there has not been much focus on (1) data-efficiency, and (2) ability to generalize across different document types and languages. In this paper, we show that when we have only a small number of labeled documents for training (~50), a straightforward transfer learning approach from a considerably structurally-different larger labeled corpus yields up to a 27 F1 point improvement over simply training on the small corpus in the target domain. We improve on this with a simple multi-domain transfer learning approach, that is currently in production use, and show that this yields up to a further 8 F1 point improvement. We make the case that data efficiency is critical to enable information extraction systems to scale to handle hundreds of different document-types, and learning good representations is critical to accomplishing this.

cs.LG↗