Abstract-- Prompting robots with natural language (NL) has largely been studied as what task to execute (goal selection, skill sequencing) rather than how to execute that task safely and efficiently in semantically rich, human-centric spaces. We address this gap with a framework that turns a large language model (LLM) into a stochastic semantic sensor whose outputs modulate a classical planner . Given a prompt and a semantic map, we draw multiple LLM "danger" judgments and apply a Bayesian bootstrap to approximate a posterior over per-class risk. Using statistics from the posterior, we create a potential cost to formulate a path planning problem. Across simulated environments and a BIM-backed digital twin, our method adapts how the robot moves in response to explicit prompts and implicit contextual information.

This study introduces a hybrid neuro-symbolic framework that achieves deterministic detection of statutory inconsistency in complex law. We use the U.S. Internal Revenue Code (IRC) as a case study because its complexity makes it a fertile domain for identifying conflicts. Our research offers a solution for detecting inconsistent provisions by combining Large Language Models (LLMs) with symbolic logic. LLM-based methods can support compliance, fairness, and statutory drafting, yet tax-specific applications remain sparse. A key challenge is that such models struggle with hierarchical processing and deep structured reasoning, especially over long text. This research addresses these gaps through experiments using GPT-4o, GPT-5, and Prolog. GPT-4o was first used to translate Section 121 into Prolog rules and refine them in SWISH. These rules were then incorporated into prompts to test whether Prolog-augmented prompting improved GPT-4o's inconsistency detection. GPT-4o, whether prompted with natural language alone or with Prolog augmentation, detected the inconsistency in only one of three strategies (33 percent accuracy), but its reasoning quality differed: natural-language prompting achieved 100 percent rule coverage, while Prolog-augmented prompting achieved 66 percent, indicating more incomplete statutory analysis. In contrast to probabilistic prompting, the hybrid Prolog model produced deterministic and reproducible results. Guided by GPT-5 for refinement, the model formalized the IRC section's competing interpretations and successfully detected an inconsistency zone. Validation tests confirm that the Prolog implementation is accurate, internally consistent, deterministic, and capable of autonomously identifying inconsistencies. These findings show that LLM-assisted formalization, anchored in symbolic logic, enables transparent and reliable statutory inconsistency detection.

Knowledge graph-based dialogue generation (KG-DG) is a challenging task requiring models to effectively incorporate external knowledge into conversational responses. While large language models (LLMs) have achieved impressive results across various NLP tasks, their ability to utilize external knowledge in KG-DG remains under-explored. We observe that LLMs often rely on internal knowledge, leading to detachment from provided knowledge graphs, even when they are given a flawlessly retrieved knowledge graph. First, we introduce LLM-KAT, an evaluation procedure for measuring knowledge attachment in generated responses. Second, we propose a simple yet effective entity anonymization technique to encourage LLMs to better leverage external knowledge. Experiments on the OpenDialKG dataset demonstrate that our approach improves LLMs' attachment on external knowledge.

Large language model agents for data analysis typically generate and execute code directly on databases. However, when applied to sensitive data, this approach poses significant security risks. To address this issue, we propose a security-motivated alternative: restrict LLMs from direct code generation and data access, and require them to interact with data exclusively through a predefined set of secure, verified tools. Although recent tool-use benchmarks exist, they primarily target tool selection and simple execution rather than the compositional, multi-step reasoning needed for complex data analysis. To reduce this gap, we introduce Indirect Data Engagement (InData), a dataset designed to assess LLMs' multi-step tool-based reasoning ability. InData includes data analysis questions at three difficulty levels--Easy, Medium, and Hard--capturing increasing reasoning complexity. We benchmark 15 open-source LLMs on InData and find that while large models (e.g., gpt-oss-120b) achieve high accuracy on Easy tasks (97.3%), performance drops sharply on Hard tasks (69.6%). These results show that current LLMs still lack robust multi-step tool-based reasoning ability. With InData, we take a step toward enabling the development and evaluation of LLMs with stronger multi-step tool-use capabilities. We will publicly release the dataset and code.

In Extended Reality (XR), rendering sound that accurately simulates real-world acoustics is pivotal in creating lifelike and believable virtual experiences. However, existing XR spatial audio rendering methods often struggle with real-time adaptation to diverse physical scenes, causing a sensory mismatch between visual and auditory cues that disrupts user immersion. To address this, we introduce SAMOSA, a novel on-device system that renders spatially accurate sound by dynamically adapting to its physical environment. SAMOSA leverages a synergistic multimodal scene representation by fusing real-time estimations of room geometry, surface materials, and semantic-driven acoustic context. This rich representation then enables efficient acoustic calibration via scene priors, allowing the system to synthesize a highly realistic Room Impulse Response (RIR). We validate our system through technical evaluation using acoustic metrics for RIR synthesis across various room configurations and sound types, alongside an expert evaluation (N=12). Evaluation results demonstrate SAMOSA's feasibility and efficacy in enhancing XR auditory realism.

Large Language Models have advanced clinical text classification, but their opaque predictions remain a critical barrier to practical adoption in research and clinical settings where investigators and physicians need to understand which parts of a patient's record drive risk signals. To address this challenge, we introduce \textbf{CALM}, short for \textbf{Classification with Additive Large Language Models}, an interpretable framework for semi-structured text where inputs are composed of semantically meaningful components, such as sections of an admission note or question-answer fields from an intake form. CALM predicts outcomes as the additive sum of each component's contribution, making these contributions part of the forward computation itself and enabling faithful explanations at both the patient and population level. The additive structure also enables clear visualizations, such as component-level risk curves similar to those used in generalized additive models, making the learned relationships easier to inspect and communicate. Although CALM expects semi-structured inputs, many clinical documents already have this form, and similar structure can often be automatically extracted from free-text notes. CALM achieves performance comparable to conventional LLM classifiers while improving trust, supporting quality-assurance checks, and revealing clinically meaningful patterns during model development and auditing.

This study aims to systematically evaluate the performance of large language models (LLMs) in abstract visual reasoning problems. We examined four LLM models (GPT-4.1-Mini, Claude-3.5-Haiku, Gemini-1.5-Flash, Llama-3.3-70b) utilizing four different reasoning architectures (single-shot, embedding-controlled repetition, self-reflection, and multi-agent) on the RAVEN-FAIR dataset. Visual responses generated through a three-stage process (JSON extraction, LLM reasoning, and Tool Function) were evaluated using SSIM and LPIPS metrics; Chain-of-Thought scores and error types (semantic hallucination, numeric misperception) were analyzed. Results demonstrate that GPT-4.1-Mini consistently achieved the highest overall accuracy across all architectures, indicating a strong reasoning capability. While the multi-agent architecture occasionally altered semantic and numeric balance across models, these effects were not uniformly beneficial. Instead, each model exhibited distinct sensitivity patterns to architectural design, underscoring that reasoning effectiveness remains model-specific. Variations in response coverage further emerged as a confounding factor that complicates direct cross-architecture comparison. To estimate the upper-bound performance of each configuration, we report the best of five independent runs, representing a best-case scenario rather than an averaged outcome. This multi-run strategy aligns with recent recommendations, which emphasize that single-run evaluations are fragile and may lead to unreliable conclusions.

Graph machine learning has advanced rapidly in tasks such as link prediction, anomaly detection, and node classification. As models scale up, pretrained graph models have become valuable intellectual assets because they encode extensive computation and domain expertise. Building on these advances, Graph Foundation Models (GFMs) mark a major step forward by jointly pretraining graph and text encoders on massive and diverse data. This unifies structural and semantic understanding, enables zero-shot inference, and supports applications such as fraud detection and biomedical analysis. However, the high pretraining cost and broad cross-domain knowledge in GFMs also make them attractive targets for model extraction attacks (MEAs). Prior work has focused only on small graph neural networks trained on a single graph, leaving the security implications for large-scale and multimodal GFMs largely unexplored. This paper presents the first systematic study of MEAs against GFMs. We formalize a black-box threat model and define six practical attack scenarios covering domain-level and graph-specific extraction goals, architectural mismatch, limited query budgets, partial node access, and training data discrepancies. To instantiate these attacks, we introduce a lightweight extraction method that trains an attacker encoder using supervised regression of graph embeddings. Even without contrastive pretraining data, this method learns an encoder that stays aligned with the victim text encoder and preserves its zero-shot inference ability on unseen graphs. Experiments on seven datasets show that the attacker can approximate the victim model using only a tiny fraction of its original training cost, with almost no loss in accuracy. These findings reveal that GFMs greatly expand the MEA surface and highlight the need for deployment-aware security defenses in large-scale graph learning systems.

Vision-language foundation models (VLMs) show promise for diverse imaging tasks but often underperform on medical benchmarks. Prior efforts to improve performance include model finetuning, which requires large domain-specific datasets and significant compute, or manual prompt engineering, which is hard to generalize and often inaccessible to medical institutions seeking to deploy these tools. These challenges motivate interest in approaches that draw on a model's embedded knowledge while abstracting away dependence on human-designed prompts to enable scalable, weight-agnostic performance improvements. To explore this, we adapt the Declarative Self-improving Python (DSPy) framework for structured automated prompt optimization in medical vision-language systems through a comprehensive, formal evaluation. We implement prompting pipelines for five medical imaging tasks across radiology, gastroenterology, and dermatology, evaluating 10 open-source VLMs with four prompt optimization techniques. Optimized pipelines achieved a median relative improvement of 53% over zero-shot prompting baselines, with the largest gains ranging from 300% to 3,400% on tasks where zero-shot performance is low. These results highlight the substantial potential of applying automated prompt optimization to medical AI systems, demonstrating significant gains for vision-based applications requiring accurate clinical image interpretation. By reducing dependence on prompt design to elicit intended outputs, these techniques allow clinicians to focus on patient care and clinical decision-making. Furthermore, our experiments offer scalability and preserve data privacy, demonstrating performance improvement on open-source VLMs. We publicly release our evaluation pipelines to support reproducible research on specialized medical tasks, available at https://github.com/DaneshjouLab/prompt-triage-lab.

Text-conditioned molecular generation aims to translate natural-language descriptions into chemical structures, enabling scientists to specify functional groups, scaffolds, and physicochemical constraints without handcrafted rules. Diffusion-based models, particularly latent diffusion models (LDMs), have recently shown promise by performing stochastic search in a continuous latent space that compactly captures molecular semantics. Yet existing methods rely on one-shot conditioning, where the entire prompt is encoded once and applied throughout diffusion, making it hard to satisfy all the requirements in the prompt. We discuss three outstanding challenges of one-shot conditioning generation, including the poor interpretability of the generated components, the failure to generate all substructures, and the overambition in considering all requirements simultaneously. We then propose three principles to address those challenges, motivated by which we propose Chain-of-Generation (CoG), a training-free multi-stage latent diffusion framework. CoG decomposes each prompt into curriculum-ordered semantic segments and progressively incorporates them as intermediate goals, guiding the denoising trajectory toward molecules that satisfy increasingly rich linguistic constraints. To reinforce semantic guidance, we further introduce a post-alignment learning phase that strengthens the correspondence between textual and molecular latent spaces. Extensive experiments on benchmark and real-world tasks demonstrate that CoG yields higher semantic alignment, diversity, and controllability than one-shot baselines, producing molecules that more faithfully reflect complex, compositional prompts while offering transparent insight into the generation process.