Yang Xie

Meng Lu, Yuxing Lu, Yuchen Zhuang et al.

Vision language models (VLMs) achieve strong performance on general image understanding but struggle to think with medical images, especially when performing multi-step reasoning through iterative visual interaction. Medical VLMs often rely on static visual embeddings and single-pass inference, preventing models from re-examining, verifying, or refining visual evidence during reasoning. While tool-integrated reasoning offers a promising path forward, open-source VLMs lack the training infrastructure to learn effective tool selection, invocation, and coordination in multi-modal medical reasoning. We introduce MedVistaGym, a scalable and interactive training environment that incentivizes tool-integrated visual reasoning for medical image analysis. MedVistaGym equips VLMs to determine when and which tools to invoke, localize task-relevant image regions, and integrate single or multiple sub-image evidence into interleaved multimodal reasoning within a unified, executable interface for agentic training. Using MedVistaGym, we train MedVistaGym-R1 to interleave tool use with agentic reasoning through trajectory sampling and end-to-end reinforcement learning. Across six medical VQA benchmarks, MedVistaGym-R1-8B exceeds comparably sized tool-augmented baselines by 19.10% to 24.21%, demonstrating that structured agentic training--not tool access alone--unlocks effective tool-integrated reasoning for medical image analysis.

Jingwei Huang, Kuroush Nezafati, Zhikai Chi et al. · gatech

Extracting structured information from clinical notes requires navigating a dense web of interdependent variables where the value of one attribute logically constrains others. Existing Large Language Model (LLM)-based extraction pipelines often struggle to capture these dependencies, leading to clinically inconsistent outputs. We propose deep reflective reasoning, a large language model agent framework that iteratively self-critiques and revises structured outputs by checking consistency among variables, the input text, and retrieved domain knowledge, stopping when outputs converge. We extensively evaluate the proposed method in three diverse oncology applications: (1) On colorectal cancer synoptic reporting from gross descriptions (n=217), reflective reasoning improved average F1 across eight categorical synoptic variables from 0.828 to 0.911 and increased mean correct rate across four numeric variables from 0.806 to 0.895; (2) On Ewing sarcoma CD99 immunostaining pattern identification (n=200), the accuracy improved from 0.870 to 0.927; (3) On lung cancer tumor staging (n=100), tumor stage accuracy improved from 0.680 to 0.833 (pT: 0.842 -> 0.884; pN: 0.885 -> 0.948). The results demonstrate that deep reflective reasoning can systematically improve the reliability of LLM-based structured data extraction under interdependence constraints, enabling more consistent machine-operable clinical datasets and facilitating knowledge discovery with machine learning and data science towards digital health.

Yang Xie, Ziqi Xu, Debo Cheng et al.

Estimating causal effects from observational data is challenging, especially in the presence of latent confounders. Much work has been done on addressing this challenge, but most of the existing research ignores the bias introduced by the post-treatment variables. In this paper, we propose a novel method of joint Variational AutoEncoder (VAE) and identifiable Variational AutoEncoder (iVAE) for learning the representations of latent confounders and latent post-treatment variables from their proxy variables, termed CPTiVAE, to achieve unbiased causal effect estimation from observational data. We further prove the identifiability in terms of the representation of latent post-treatment variables. Extensive experiments on synthetic and semi-synthetic datasets demonstrate that the CPTiVAE outperforms the state-of-the-art methods in the presence of latent confounders and post-treatment variables. We further apply CPTiVAE to a real-world dataset to show its potential application.

Meng Lu, Ran Xu, Yi Fang et al.

While recent vision-language models (VLMs) demonstrate strong image understanding, their ability to "think with images", i.e., to reason through multi-step visual interactions, remains limited. We introduce VISTA-Gym, a scalable training environment for incentivizing tool-integrated visual reasoning capabilities in VLMs. VISTA-Gym unifies diverse real-world multimodal reasoning tasks (7 tasks from 13 datasets in total) with a standardized interface for visual tools (e.g., grounding, parsing), executable interaction loops, verifiable feedback signals, and efficient trajectory logging, enabling visual agentic reinforcement learning at scale. While recent VLMs exhibit strong text-only reasoning, both proprietary and open-source models still struggle with tool selection, invocation, and coordination. With VISTA-Gym, we train VISTA-R1 to interleave tool-use with agentic reasoning via multi-turn trajectory sampling and end-to-end reinforcement learning. Extensive experiments across 11 public reasoning-intensive VQA benchmarks show that VISTA-R1-8B outperforms state-of-the-art baselines with similar sizes by 9.51%-18.72%, demonstrating VISTA-Gym as an effective training ground to unlock the tool-integrated reasoning capabilities for VLMs.

Ran Xu, Yuchen Zhuang, Yishan Zhong et al. · gatech

We introduce MedAgentGym, a scalable and interactive training environment designed to enhance coding-based biomedical reasoning capabilities in large language model (LLM) agents. MedAgentGym comprises 72,413 task instances across 129 categories derived from 12 authentic real-world biomedical scenarios. Tasks are encapsulated within executable sandbox environments, each featuring detailed task specifications, interactive feedback mechanisms, verifiable ground truth annotations, and scalable training trajectory generation. Extensive benchmarking of 29 LLMs reveals substantial performance disparities in biomedical data science between commercial and open-source LLMs. Leveraging efficient multi-threaded and multi-turn trajectory sampling in MedAgentGym, Med-Copilot achieves performance gains of +43.02% and +45.28% from offline and online reinforcement learning, respectively, demonstrating MedAgentGym as an effective training ground while establishing itself as a cost-effective, privacy-preserving alternative competitive with proprietary LLMs (gpt-4o). By offering a unified execution environment with a comprehensive benchmark and accessible, extensible training resources, MedAgentGym delivers an integrated platform to develop LLM-based coding assistants for advanced biomedical data science.

Jingwei Huang, Kuroush Nezafati, Ismael Villanueva-Miranda et al.

This study introduces a LLMs powered multiagent ensemble method to address challenges in hallucination and data labeling, particularly in large-scale EHR datasets. Manual labeling of such datasets requires domain expertise and is labor-intensive, time-consuming, expensive, and error-prone. To overcome this bottleneck, we developed an ensemble LLMs method and demonstrated its effectiveness in two real-world tasks: (1) labeling a large-scale unlabeled ECG dataset in MIMIC-IV; (2) identifying social determinants of health (SDOH) from the clinical notes of EHR. Trading off benefits and cost, we selected a pool of diverse open source LLMs with satisfactory performance. We treat each LLM's prediction as a vote and apply a mechanism of majority voting with minimal winning threshold for ensemble. We implemented an ensemble LLMs application for EHR data labeling tasks. By using the ensemble LLMs and natural language processing, we labeled MIMIC-IV ECG dataset of 623,566 ECG reports with an estimated accuracy of 98.2%. We applied the ensemble LLMs method to identify SDOH from social history sections of 1,405 EHR clinical notes, also achieving competitive performance. Our experiments show that the ensemble LLMs can outperform individual LLM even the best commercial one, and the method reduces hallucination errors. From the research, we found that (1) the ensemble LLMs method significantly reduces the time and effort required for labeling large-scale EHR data, automating the process with high accuracy and quality; (2) the method generalizes well to other text data labeling tasks, as shown by its application to SDOH identification; (3) the ensemble of a group of diverse LLMs can outperform or match the performance of the best individual LLM; and (4) the ensemble method substantially reduces hallucination errors. This approach provides a scalable and efficient solution to data-labeling challenges.

Debo Cheng, Yang Xie, Ziqi Xu et al.

In causal inference, it is a fundamental task to estimate the causal effect from observational data. However, latent confounders pose major challenges in causal inference in observational data, for example, confounding bias and M-bias. Recent data-driven causal effect estimators tackle the confounding bias problem via balanced representation learning, but assume no M-bias in the system, thus they fail to handle the M-bias. In this paper, we identify a challenging and unsolved problem caused by a variable that leads to confounding bias and M-bias simultaneously. To address this problem with co-occurring M-bias and confounding bias, we propose a novel Disentangled Latent Representation learning framework for learning latent representations from proxy variables for unbiased Causal effect Estimation (DLRCE) from observational data. Specifically, DLRCE learns three sets of latent representations from the measured proxy variables to adjust for the confounding bias and M-bias. Extensive experiments on both synthetic and three real-world datasets demonstrate that DLRCE significantly outperforms the state-of-the-art estimators in the case of the presence of both confounding bias and M-bias.

Mulong Xie, Yang Xie

Chat-based natural language interfaces have emerged as the dominant paradigm for human-agent interaction, yet they fundamentally constrain engagement with structured information and complex tasks. We identify three inherent limitations: the mismatch between structured data and linear text, the high entropy of unconstrained natural language input, and the lack of persistent, evolving interaction state. We introduce Software as Content (SaC), a paradigm in which dynamically generated agentic applications serve as the primary medium of human-agent interaction. Rather than communicating through sequential text exchange, this medium renders task-specific interfaces that present structured information and expose actionable affordances through which users iteratively guide agent behavior without relying solely on language. These interfaces persist and evolve across interaction cycles, transforming from transient responses into a shared, stateful interaction layer that progressively converges toward personalized, task-specific software. We formalize SaC through a human-agent-environment interaction model, derive design principles for generating and evolving agentic applications, and present a system architecture that operationalizes the paradigm. We evaluate across representative tasks of selection, exploration, and execution, demonstrating technical viability and expressive range, while identifying boundary conditions under which natural language remains preferable. By reframing interfaces as dynamically generated software artifacts, SaC opens a new design space for human-AI interaction, positioning dynamic software as a concrete and tractable research object.

Peiran Quan, Zifan Gu, Zhuo Zhao et al.

Foundation models (FMs) have transformed computational pathology by providing powerful, general-purpose feature extractors. However, adapting and benchmarking individual FMs for specific diagnostic tasks is often time-consuming and resource-intensive, especially given their scale and diversity. To address this challenge, we introduce Group-Aggregative Selection Multi-Instance Learning (GAS-MIL), a flexible ensemble framework that seamlessly integrates features from multiple FMs, preserving their complementary strengths without requiring manual feature selection or extensive task-specific fine-tuning. Across classification tasks in three cancer datasets-prostate (PANDA), ovarian (UBC-OCEAN), and breast (TCGA-BrCa)-GAS-MIL consistently achieves superior or on-par performance relative to individual FMs and established MIL methods, demonstrating its robustness and generalizability. By enabling efficient integration of heterogeneous FMs, GAS-MIL streamlines model deployment for pathology and provides a scalable foundation for future multimodal and precision oncology applications.

Wenting Ye, Chen Li, Yang Xie et al.

Identifying and discovering drug-target interactions(DTIs) are vital steps in drug discovery and development. They play a crucial role in assisting scientists in finding new drugs and accelerating the drug development process. Recently, knowledge graph and knowledge graph embedding (KGE) models have made rapid advancements and demonstrated impressive performance in drug discovery. However, such models lack authenticity and accuracy in drug target identification, leading to an increased misjudgment rate and reduced drug development efficiency. To address these issues, we focus on the problem of drug-target interactions, with knowledge mapping as the core technology. Specifically, a causal intervention-based confidence measure is employed to assess the triplet score to improve the accuracy of the drug-target interaction prediction model. Experimental results demonstrate that the developed confidence measurement method based on causal intervention can significantly enhance the accuracy of DTI link prediction, particularly for high-precision models. The predicted results are more valuable in guiding the design and development of subsequent drug development experiments, thereby significantly improving the efficiency of drug development.

Olawale Onabola, Zhuang Ma, Yang Xie et al.

Subtle and overt racism is still present both in physical and online communities today and has impacted many lives in different segments of the society. In this short piece of work, we present how we're tackling this societal issue with Natural Language Processing. We are releasing BiasCorp, a dataset containing 139,090 comments and news segment from three specific sources - Fox News, BreitbartNews and YouTube. The first batch (45,000 manually annotated) is ready for publication. We are currently in the final phase of manually labeling the remaining dataset using Amazon Mechanical Turk. BERT has been used widely in several downstream tasks. In this work, we present hBERT, where we modify certain layers of the pretrained BERT model with the new Hopfield Layer. hBert generalizes well across different distributions with the added advantage of a reduced model complexity. We are also releasing a JavaScript library and a Chrome Extension Application, to help developers make use of our trained model in web applications (say chat application) and for users to identify and report racially biased contents on the web respectively.

Yuntao Liu, Michael Zuzak, Yang Xie et al.

Logic locking is a hardware security technique to intellectual property (IP) against security threats in the IC supply chain, especially untrusted fabs. Such techniques incorporate additional locking circuitry within an IC that induces incorrect functionality when an incorrect key is provided. The amount of error induced is known as the effectiveness of the locking technique. "SAT attacks" provide a strong mathematical formulation to find the correct key of locked circuits. In order to achieve high SAT resilience(i.e. complexity of SAT attacks), many conventional logic locking schemes fail to inject sufficient error into the circuit. For example, in the case of SARLock and Anti-SAT, there are usually very few (or only one) input minterms that cause any error at the circuit output. The state-of-the-art stripped functionality logic locking (SFLL) technique introduced a trade-off between SAT resilience and effectiveness. In this work, we prove that such a trade-off is universal in logic locking. In order to attain high effectiveness of locking without compromising SAT resilience, we propose a novel logic locking scheme, called Strong Anti-SAT (SAS). In addition to SAT attacks, removal-based attacks are also popular against logic locking. Based on SAS, we propose Robust SAS (RSAS) which is resilient to removal attacks and maintains the same SAT resilience and as effectiveness as SAS. SAS and RSAS have the following significant improvements over existing techniques. (1) SAT resilience of SAS and RSAS against SAT attack is not compromised by increase in effectiveness. (2) In contrast to prior work focusing solely on the circuit-level locking impact, we integrate SAS-locked modules into a processor and show that SAS has a high application-level impact. (3) Our experiments show that SAS and RSAS exhibit better SAT resilience than SFLL and have similar effectiveness.

Yang Xie, Sijjad Ali Khuhro, Fuyou Miao et al.

Traditional threshold secret sharing cannot realizing all access structures of secret sharing. So, Ito introduced the concept of Secret sharing scheme realizing general access structure. But Its scheme has to send multiple shares to each trustee. In this paper, we proposed two new secret sharing schemes realizing general access structures by only assigning one share to each trustee. Our proposed second scheme is a perfect secret sharing scheme. Furthermore, our schemes can realize any access structures.

Shidan Wang, Tao Wang, Lin Yang et al.

The spatial distributions of different types of cells could reveal a cancer cell growth pattern, its relationships with the tumor microenvironment and the immune response of the body, all of which represent key hallmarks of cancer. However, manually recognizing and localizing all the cells in pathology slides are almost impossible. In this study, we developed an automated cell type classification pipeline, ConvPath, which includes nuclei segmentation, convolutional neural network-based tumor, stromal and lymphocytes classification, and extraction of tumor microenvironment related features for lung cancer pathology images. The overall classification accuracy is 92.9% and 90.1% in training and independent testing datasets, respectively. By identifying cells and classifying cell types, this pipeline can convert a pathology image into a spatial map of tumor, stromal and lymphocyte cells. From this spatial map, we can extracted features that characterize the tumor micro-environment. Based on these features, we developed an image feature-based prognostic model and validated the model in two independent cohorts. The predicted risk group serves as an independent prognostic factor, after adjusting for clinical variables that include age, gender, smoking status, and stage.

Yuntao Liu, Yang Xie, Ankur Srivastava

While neural networks demonstrate stronger capabilities in pattern recognition nowadays, they are also becoming larger and deeper. As a result, the effort needed to train a network also increases dramatically. In many cases, it is more practical to use a neural network intellectual property (IP) that an IP vendor has already trained. As we do not know about the training process, there can be security threats in the neural IP: the IP vendor (attacker) may embed hidden malicious functionality, i.e. neural Trojans, into the neural IP. We show that this is an effective attack and provide three mitigation techniques: input anomaly detection, re-training, and input preprocessing. All the techniques are proven effective. The input anomaly detection approach is able to detect 99.8% of Trojan triggers although with 12.2% false positive. The re-training approach is able to prevent 94.1% of Trojan triggers from triggering the Trojan although it requires that the neural IP be reconfigurable. In the input preprocessing approach, 90.2% of Trojan triggers are rendered ineffective and no assumption about the neural IP is needed.