CoAdapt-GUI: Joint Workflow Context and Policy Adaptation for Unseen GUI Applications

Mobile GUI agents remain brittle when deployed to applications absent from source training. We study novel-app generalization under a limited target interaction budget and without target demonstrations. We introduce CoAdapt-GUI, a test-time adaptation (TTA) framework that jointly adapts structured workflow context and policy from the agent's own target-app rollouts and rewards. The workflow context retains transferable procedures, failure modes, and verification rules while excluding app-bound source details. This separation allows reusable workflow knowledge to guide adaptation without transferring source-interface state. For policy adaptation, task-context-matched group-relative optimization updates a LoRA adapter on a frozen vision-language model. Across two unseen-app evaluations, CoAdapt-GUI reaches 45.0% on AndroidWorld-Generalization, compared with 37.5% for the reported Policy-Only TTA baseline, and raises AndroidWorld Plus performance from 38.6% to 52.9%. These results show that transfer-constrained workflow context provides substantial gains and that joint policy adaptation further improves held-out performance.

StepReflect: Structured UI Transition Reflection for Mobile GUI Agents

Autonomous mobile GUI agents require accurate action reflection for reliable long-horizon execution. Existing approaches rely on open-ended multimodal reasoning after each action, which is costly and poorly matched to the structured nature of GUI state transitions. We propose StepReflect, which formulates per-step GUI reflection as supervised structured prediction conditioned on explicit transition specifications and paired visual evidence. StepReflect is trained through a staged pipeline combining supervised fine-tuning, teacher-student distillation, and preference- and reward-based refinement. Offline, the resulting 8B model achieves 82.16% transition-level accuracy on AndroidWorld, exceeding zero-shot GPT-5.2 by 11.83 percentage points under the same structured input. Online, across M3A, Agent-SAMA, MAI-UI-8B, and Seed-2.0-Pro, StepReflect achieves higher task success in three of four agent configurations and remains within one successful task of the GPT-5.2 Reflection Agent in the fourth. It also reduces paid API charges relative to GPT-based reflection in all four configurations. These results establish StepReflect as a practical, locally deployable alternative to repeated frontier-model reflection for long-horizon mobile GUI agents.

End-to-End Fairness Optimization with Fair Decision-Focused Learning

Many real-world systems rely on predictive models to inform decisions, and fairness concerns arise in both the prediction and decision stages. We introduce end-to-end fairness optimization (E2EFO) as a unifying framework that integrates fairness across the prediction-to-decision pipeline. We focus on resource allocation with group-based fairness: the prediction task estimates allocation impacts while limiting accuracy disparity across groups, and the decision task distributes those impacts equitably by optimizing a group-based alpha-fairness measure. Within this framework, we propose fair decision-focused learning (FDFL), a training paradigm that jointly accounts for prediction accuracy, prediction fairness, and decision regret -- the loss in decision fairness due to imperfect predictions. FDFL trains the predictor by gradient descent, combining the objective gradients through multi-task learning techniques. The core computational challenge is the decision Jacobian with respect to the predictor parameters: we derive exact closed-form formulas for a tractable class of fair allocation and apply a differentiable optimization layer in the general case. We further establish a finite-sample generalization bound for the scalarized FDFL objective. Numerical experiments on a healthcare-based single resource allocation and a synthetic multiple resource allocation illustrate the value of jointly accounting for prediction fairness and decision fairness in prediction-informed decision-making.

Deployment-Time Memorization in Foundation-Model Agents

Foundation-model agents are increasingly long-lived systems that remember users across interactions, making memorization an explicit deployment-time function rather than solely a property of model weights. Existing work addresses parametric memorization or audits fixed memory configurations, but does not characterize how memory-design choices jointly shape personalization utility, extraction risk, and deletion fidelity. We study this surface as deployment-time memorization, formulating agent memory as a privacy-utility frontier measured by Personalization Recall (PR) and Adversarial Extraction Rate (AER), and sweeping three memory-design knobs: summarization aggressiveness, retrieval breadth (k), and deletion mode. We further introduce the Forgetting Residue Score (FRS) to quantify whether deleted information remains recoverable from derived memory tiers. On LongMemEval, key-fact summarization reduces canary extraction by 76% on Gemma 3 12B and 64% on GPT-4o-mini while preserving nearly all personalization recall; critically, once content is compressed away, increasing k no longer restores leakage. The same compression, however, induces a deletion-fidelity failure: raw-only deletion leaves derived summary copies recoverable in approximately 20% of instances, and only full-pipeline purge or tombstone redaction drives worst-tier residue to zero. Together, these results establish that persistent agent memory must be evaluated as a first-class memorization mechanism -- assessed by what it helps agents recall, what it makes extractable, and what it can truly erase.

RealICU: Do LLM Agents Understand Long-Context ICU Data? A Benchmark Beyond Behavior Imitation

Intensive care units (ICU) generate long, dense and evolving streams of clinical information, where physicians must repeatedly reassess patient states under time pressure, underscoring a clear need for reliable AI decision support. Existing ICU benchmarks typically treat historical clinician actions as ground truth. However, these actions are made under incomplete information and limited temporal context of the underlying patient state, and may therefore be suboptimal, making it difficult to assess the true reasoning capabilities of AI systems. We introduce RealICU, a hindsight-annotated benchmark for evaluating large language models (LLMs) under realistic ICU conditions, where labels are created after senior physicians review the full patient trajectory. We formulate four physician-motivated tasks: assess Patient Status, Acute Problems, Recommended Actions, and Red Flag actions that risk unsafe outcomes. We partition each trajectory with 30-min windows and release two datasets: RealICU-Gold with 930-window annotations from 94 MIMIC-IV patients, and RealICU-Scale with 11,862 windows extended by Oracle, a physician-validated LLM hindsight labeler. Existing LLMs including memory-augmented ones performed poorly on RealICU, exposing two failure modes: a recall-safety tradeoff for clinical recommendations, and an anchoring bias to early interpretations of the patient. We further introduce ICU-Evo to study structured-memory agents that improves long-horizon reasoning but does not fully eliminate safety failures. Together, RealICU provides a clinically grounded testbed for measuring and improving AI sequential decision-support in high-stakes care. Project page: https://chengzhi-leo.github.io/RealICU-Bench/

IPI-proxy: An Intercepting Proxy for Red-Teaming Web-Browsing AI Agents Against Indirect Prompt Injection

Web-browsing AI agents are increasingly deployed in enterprise settings under strict whitelists of approved domains, yet adversaries can still influence them by embedding hidden instructions in the HTML pages those domains serve. Existing red-teaming resources fall short of this scenario: prompt-injection benchmarks ship pre-built adversarial pages that whitelisted agents cannot reach, and generic LLM scanners probe the model API rather than its retrieved content. We present IPI-proxy, an open-source toolkit for red-teaming web-browsing agents against indirect prompt injection (IPI). At its core is an intercepting proxy that rewrites real HTTP responses from whitelisted domains in flight, embedding payloads drawn from a unified library of 820 deduplicated attack strings extracted from six published benchmarks (BIPIA, InjecAgent, AgentDojo, Tensor Trust, WASP, and LLMail-Inject). A YAML-driven test harness independently parameterizes the payload set, the embedding technique (HTML comment, invisible CSS, or LLM-generated semantic prose), and the HTML insertion point (6 locations from \icode{head\_meta} to \icode{script\_comment}), enabling parameter-sweep evaluation without mock pages or sandboxed environments. A companion exfiltration tracker logs successful callbacks. This paper describes the threat model, situates IPI-proxy among contemporary IPI benchmarks and red-teaming tools, and details its architecture, design decisions, and configuration interface. By bridging static benchmarks and live deployment, IPI-proxy gives AI security teams a reproducible substrate for measuring and hardening web-browsing agents against indirect prompt injection on the same retrieval surface attackers exploit in production.

Dialogue Act Patterns in GenAI-Mediated L2 Oral Practice: A Sequential Analysis of Learner-Chatbot Interactions

While generative AI (GenAI) voice chatbots offer scalable opportunities for second language (L2) oral practice, the interactional processes related to learners' gains remain underexplored. This study investigates dialogue act (DA) patterns in interactions between Grade 9 Chinese English as a foreign language (EFL) learners and a GenAI voice chatbot over a 10-week intervention. Seventy sessions from 12 students were annotated by human coders using a pedagogy-informed coding scheme, yielding 6,957 coded DAs. DA distributions and sequential patterns were compared between high- and low-progress sessions. At the DA level, high-progress sessions showed more learner-initiated questions, whereas low-progress sessions exhibited higher rates of clarification-seeking, indicating greater comprehension difficulty. At the sequential level, high-progress sessions were characterised by more frequent prompting-based corrective feedback sequences, consistently positioned after learner responses, highlighting the role of feedback type and timing in effective interaction. Overall, these findings underscore the value of a dialogic lens in GenAI chatbot design, contribute a pedagogy-informed DA coding framework, and inform the design of adaptive GenAI chatbots for L2 education.

Towards Structured, State-Aware, and Execution-Grounded Reasoning for Software Engineering Agents

Software Engineering (SE) agents have shown promising abilities in supporting various SE tasks. Current SE agents remain fundamentally reactive, making decisions mainly based on conversation history and the most recent response. However, this reactive design provides no explicit structure or persistent state within the agent's memory, making long-horizon reasoning challenging. As a result, SE agents struggle to maintain a coherent understanding across reasoning steps, adapt their hypotheses as new evidence emerges, or incorporate execution feedback into the mental reasoning model of the system state. In this position paper, we argue that, to further advance SE agents, we need to move beyond reactive behavior toward a structured, state-aware, and execution-grounded reasoning. We outline how explicit structure, persistent and evolving state, and the integration of execution-grounded feedback can help SE agents perform more coherent and reliable reasoning in long-horizon tasks. We also provide an initial roadmap for developing next-generation SE agents that can more effectively perform real-world tasks.

Does Privacy Always Harm Fairness? Data-Dependent Trade-offs via Chernoff Information Neural Estimation

Fairness and privacy are two vital pillars of trustworthy machine learning. Despite extensive research on these individual topics, the relationship between fairness and privacy has received significantly less attention. In this paper, we utilize the information-theoretic measure Chernoff Information to highlight the data-dependent nature of the relationship among the triad of fairness, privacy, and accuracy. We first define Noisy Chernoff Difference, a tool that allows us to analyze the relationship among the triad simultaneously. We then show that for synthetic data, this value behaves in 3 distinct ways (depending on the distribution of the data). We highlight the data distributions involved in these cases and explore their fairness and privacy implications. Additionally, we show that Noisy Chernoff Difference acts as a proxy for the steepness of the fairness-accuracy curves. Finally, we propose a method for estimating Chernoff Information on data from unknown distributions and utilize this framework to examine the triad dynamic on real datasets. This work builds towards a unified understanding of the fairness-privacy-accuracy relationship and highlights its data-dependent nature.

Diffusion Large Language Models for Black-Box Optimization

Offline black-box optimization (BBO) aims to find optimal designs based solely on an offline dataset of designs and their labels. Such scenarios frequently arise in domains like DNA sequence design and robotics, where only a few labeled data points are available. Traditional methods typically rely on task-specific proxy or generative models, overlooking the in-context learning capabilities of pre-trained large language models (LLMs). Recent efforts have adapted autoregressive LLMs to BBO by framing task descriptions and offline datasets as natural language prompts, enabling direct design generation. However, these designs often contain bidirectional dependencies, which left-to-right models struggle to capture. In this paper, we explore diffusion LLMs for BBO, leveraging their bidirectional modeling and iterative refinement capabilities. This motivates our in-context denoising module: we condition the diffusion LLM on the task description and the offline dataset, both formatted in natural language, and prompt it to denoise masked designs into improved candidates. To guide the generation toward high-performing designs, we introduce masked diffusion tree search, which casts the denoising process as a step-wise Monte Carlo Tree Search that dynamically balances exploration and exploitation. Each node represents a partially masked design, each denoising step is an action, and candidates are evaluated via expected improvement under a Gaussian Process trained on the offline dataset. Our method, dLLM, achieves state-of-the-art results in few-shot settings on design-bench.