Publications

@misc{wang_gameplayqa_2026,

title = {GameplayQA: A Benchmarking Framework for Decision-Dense POV-Synced Multi-Video Understanding of 3D Virtual Agents},

author = {Yunzhe Wang and Runhui Xu and Kexin Zheng and Tianyi Zhang and Jayavibhav Niranjan Kogundi and Soham Hans and Volkan Ustun},

url = {http://arxiv.org/abs/2603.24329},

doi = {10.48550/arXiv.2603.24329},

year  = {2026},

date = {2026-04-01},

urldate = {2026-06-18},

publisher = {arXiv},

abstract = {Multimodal LLMs are increasingly deployed as perceptual backbones for autonomous agents in 3D environments, from robotics to virtual worlds. These applications require agents to perceive rapid state changes, attribute actions to the correct entities, and reason about concurrent multi-agent behaviors from a first-person perspective, capabilities that existing benchmarks do not adequately evaluate. We introduce GameplayQA, a framework for evaluating agentic-centric perception and reasoning through video understanding. Specifically, we densely annotate multiplayer 3D gameplay videos at 1.22 labels/second, with time-synced, concurrent captions of states, actions, and events structured around a triadic system of Self, Other Agents, and the World, a natural decomposition for multi-agent environments. From these annotations, we refined 2.4K diagnostic QA pairs organized into three levels of cognitive complexity, accompanied by a structured distractor taxonomy that enables fine-grained analysis of where models hallucinate. Evaluation of frontier MLLMs reveals a substantial gap from human performance, with common failures in temporal and cross-video grounding, agent-role attribution, and handling the decision density of the game. We hope GameplayQA stimulates future research at the intersection of embodied AI, agentic perception, and world modeling.},

note = {arXiv:2603.24329 [cs.CL]},

keywords = {AI, DTIC, LLM, UARC, Virtual Agents},

pubstate = {published},

tppubtype = {misc}

}

@misc{feng_howtobench_2026,

title = {HoWToBench: Holistic Evaluation for LLM's Capability in Human-level Writing using Tree of Writing},

author = {Andrew Zhuoer Feng and Cunxiang Wang and Yu Luo and Lin Fan and Yilin Zhou and Zikang Wang and Xiaotao Gu and Jie Tang and Hongning Wang and Minlie Huang},

url = {http://arxiv.org/abs/2604.19071},

doi = {10.48550/arXiv.2604.19071},

year  = {2026},

date = {2026-04-01},

urldate = {2026-06-17},

publisher = {arXiv},

abstract = {Evaluating the writing capabilities of large language models (LLMs) remains a significant challenge due to the multidimensional nature of writing skills and the limitations of existing metrics. LLM's performance in thousand-words level and open-ended writing is inadequately assessed by traditional reference-based metrics or modern LLM-as-a-judge methods. We propose Tree-of-Writing (ToW), to resolve the implicit inconsistency often found when LLM-as-a-judge aggregates all sub-features in text evaluation. ToW incorporates a tree-structured workflow by explicitly modeling the aggregation weights of sub-features. We also present HowToBench, a large-scale Chinese writing benchmark encompassing 12 genres and 1302 instructions across three task categories: contextual completion, outline-guided writing, and open-ended generation. ToW successfully mitigates the biases, achieving a 0.93 Pearson correlation with human judgments. Furthermore, we detect that both overlap-based text generation metrics and popular LLM-as-a-judge practices are vulnerable to textual disturbances, while ToW is robust to them. We also uncover a negative correlation between input length and content-related scores in the Guide task, showcasing that it cannot be simply improved by input-side information piling.},

note = {arXiv:2604.19071 [cs.CL]},

keywords = {LLM},

pubstate = {published},

tppubtype = {misc}

}

@misc{kwon_can_2026,

title = {Can LLMs Truly Embody Human Personality? Analyzing AI and Human Behavior Alignment in Dispute Resolution},

author = {Deuksin Kwon and Kaleen Shrestha and Bin Han and Spencer Lin and James Hale and Jonathan Gratch and Maja Matarić and Gale M. Lucas},

url = {http://arxiv.org/abs/2602.07414},

doi = {10.48550/arXiv.2602.07414},

year  = {2026},

date = {2026-02-01},

urldate = {2026-06-18},

publisher = {arXiv},

abstract = {Large language models (LLMs) are increasingly used to simulate human behavior in social settings such as legal mediation, negotiation, and dispute resolution. However, it remains unclear whether these simulations reproduce the personality-behavior patterns observed in humans. Human personality, for instance, shapes how individuals navigate social interactions, including strategic choices and behaviors in emotionally charged interactions. This raises the question: Can LLMs, when prompted with personality traits, reproduce personality-driven differences in human conflict behavior? To explore this, we introduce an evaluation framework that enables direct comparison of human-human and LLM-LLM behaviors in dispute resolution dialogues with respect to Big Five Inventory (BFI) personality traits. This framework provides a set of interpretable metrics related to strategic behavior and conflict outcomes. We additionally contribute a novel dataset creation methodology for LLM dispute resolution dialogues with matched scenarios and personality traits with respect to human conversations. Finally, we demonstrate the use of our evaluation framework with three contemporary closed-source LLMs and show significant divergences in how personality manifests in conflict across different LLMs compared to human data, challenging the assumption that personality-prompted agents can serve as reliable behavioral proxies in socially impactful applications. Our work highlights the need for psychological grounding and validation in AI simulations before real-world use.},

note = {arXiv:2602.07414 [cs.AI]},

keywords = {ARO, DTIC, LLM, Simulation, UARC},

pubstate = {published},

tppubtype = {misc}

}

@misc{han_personality_2026,

title = {Personality Expression Across Contexts: Linguistic and Behavioral Variation in LLM Agents},

author = {Bin Han and Deuksin Kwon and Jonathan Gratch},

url = {http://arxiv.org/abs/2602.01063},

doi = {10.48550/arXiv.2602.01063},

year  = {2026},

date = {2026-02-01},

urldate = {2026-06-19},

publisher = {arXiv},

abstract = {Large Language Models (LLMs) can be conditioned with explicit personality prompts, yet their behavioral realization often varies depending on context. This study examines how identical personality prompts lead to distinct linguistic, behavioral, and emotional outcomes across four conversational settings: ice-breaking, negotiation, group decision, and empathy tasks. Results show that contextual cues systematically influence both personality expression and emotional tone, suggesting that the same traits are expressed differently depending on social and affective demands. This raises an important question for LLM-based dialogue agents: whether such variations reflect inconsistency or context-sensitive adaptation akin to human behavior. Viewed through the lens of Whole Trait Theory, these findings highlight that LLMs exhibit context-sensitive rather than fixed personality expression, adapting flexibly to social interaction goals and affective conditions.},

note = {arXiv:2602.01063 [cs.CL]},

keywords = {LLM, Virtual Agents},

pubstate = {published},

tppubtype = {misc}

}

@misc{tak_sparks_2026,

title = {Sparks of Rationality: Do Reasoning LLMs Align with Human Judgment and Choice?},

author = {Ala N. Tak and Amin Banayeeanzade and Anahita Bolourani and Fatemeh Bahrani and Ashutosh Chaubey and Sai Praneeth Karimireddy and Norbert Schwarz and Jonathan Gratch},

url = {http://arxiv.org/abs/2601.22329},

doi = {10.48550/arXiv.2601.22329},

year  = {2026},

date = {2026-01-01},

urldate = {2026-06-19},

publisher = {arXiv},

abstract = {Large Language Models (LLMs) are increasingly positioned as decision engines for hiring, healthcare, and economic judgment, yet real-world human judgment reflects a balance between rational deliberation and emotion-driven bias. If LLMs are to participate in high-stakes decisions or serve as models of human behavior, it is critical to assess whether they exhibit analogous patterns of (ir)rationalities and biases. To this end, we evaluate multiple LLM families on (i) benchmarks testing core axioms of rational choice and (ii) classic decision domains from behavioral economics and social norms where emotions are known to shape judgment and choice. Across settings, we show that deliberate "thinking" reliably improves rationality and pushes models toward expected-value maximization. To probe human-like affective distortions and their interaction with reasoning, we use two emotion-steering methods: in-context priming (ICP) and representation-level steering (RLS). ICP induces strong directional shifts that are often extreme and difficult to calibrate, whereas RLS produces more psychologically plausible patterns but with lower reliability. Our results suggest that the same mechanisms that improve rationality also amplify sensitivity to affective interventions, and that different steering methods trade off controllability against human-aligned behavior. Overall, this points to a tension between reasoning and affective steering, with implications for both human simulation and the safe deployment of LLM-based decision systems.},

note = {arXiv:2601.22329 [cs.AI]},

keywords = {Emotions, LLM},

pubstate = {published},

tppubtype = {misc}

}

@misc{xu_optimizing_2025,

title = {Optimizing Federated Learning in the Era of LLMs: Message Quantization and Streaming},

author = {Ziyue Xu and Zhihong Zhang and Holger R. Roth and Chester Chen and Yan Cheng and Andrew Feng},

url = {http://arxiv.org/abs/2511.16450},

doi = {10.48550/arXiv.2511.16450},

year  = {2025},

date = {2025-11-01},

urldate = {2026-06-25},

publisher = {arXiv},

abstract = {Federated Learning (FL) offers a promising solution for training machine learning models across distributed data sources while preserving data privacy. However, FL faces critical challenges related to communication overhead and local resource constraints, especially in the era of Large Language Models (LLMs) with billions of parameters. The sheer size of these models exacerbates both memory and communication constraints, making efficient transmission and processing essential for practical deployment. NVIDIA FLARE, an open-source SDK for federated learning, addresses these challenges by introducing advanced communication capabilities. Building upon existing solutions for large object streaming, we enhance FL workflows for LLMs through two key techniques: message quantization and container/file streaming. Quantization reduces message size, while streaming enables efficient memory management, improving scalability and integration with existing workflows. These advancements significantly enhance the robustness and efficiency of FL with LLMs, ensuring better performance in real-world federated learning scenarios.},

note = {arXiv:2511.16450 [cs.DC]},

keywords = {LLM, Machine Learning},

pubstate = {published},

tppubtype = {misc}

}

@misc{hans_towards_2025,

title = {Towards AI-Assisted Generation of Military Training Scenarios},

author = {Soham Hans and Volkan Ustun and Benjamin Nye and James Sterrett and Matthew Green},

url = {http://arxiv.org/abs/2511.07690},

doi = {10.48550/arXiv.2511.07690},

year  = {2025},

date = {2025-11-01},

urldate = {2026-06-17},

publisher = {arXiv},

abstract = {Achieving expert-level performance in simulation-based training relies on the creation of complex, adaptable scenarios, a traditionally laborious and resource intensive process. Although prior research explored scenario generation for military training, pre-LLM AI tools struggled to generate sufficiently complex or adaptable scenarios. This paper introduces a multi-agent, multi-modal reasoning framework that leverages Large Language Models (LLMs) to generate critical training artifacts, such as Operations Orders (OPORDs). We structure our framework by decomposing scenario generation into a hierarchy of subproblems, and for each one, defining the role of the AI tool: (1) generating options for a human author to select from, (2) producing a candidate product for human approval or modification, or (3) generating textual artifacts fully automatically. Our framework employs specialized LLM-based agents to address distinct subproblems. Each agent receives input from preceding subproblem agents, integrating both text-based scenario details and visual information (e.g., map features, unit positions and applies specialized reasoning to produce appropriate outputs. Subsequent agents process these outputs sequentially, preserving logical consistency and ensuring accurate document generation. This multi-agent strategy overcomes the limitations of basic prompting or single-agent approaches when tackling such highly complex tasks. We validate our framework through a proof-of-concept that generates the scheme of maneuver and movement section of an OPORD while estimating map positions and movements as a precursor demonstrating its feasibility and accuracy. Our results demonstrate the potential of LLM-driven multi-agent systems to generate coherent, nuanced documents and adapt dynamically to changing conditions, advancing automation in scenario generation for military training.},

note = {arXiv:2511.07690 [cs.AI]},

keywords = {AI, DTIC, Learning Sciences, LLM, UARC},

pubstate = {published},

tppubtype = {misc}

}

@misc{zhang_individualized_2025,

title = {Individualized Cognitive Simulation in Large Language Models: Evaluating Different Cognitive Representation Methods},

author = {Tianyi Zhang and Xiaolin Zhou and Yunzhe Wang and Erik Cambria and David Traum and Rui Mao},

url = {http://arxiv.org/abs/2510.20252},

doi = {10.48550/arXiv.2510.20252},

year  = {2025},

date = {2025-10-01},

urldate = {2026-07-06},

publisher = {arXiv},

abstract = {Individualized cognitive simulation (ICS) aims to build computational models that approximate the thought processes of specific individuals. While large language models (LLMs) convincingly mimic surface-level human behavior such as role-play, their ability to simulate deeper individualized cognitive processes remains poorly understood. To address this gap, we introduce a novel task that evaluates different cognitive representation methods in ICS. We construct a dataset from recently published novels (later than the release date of the tested LLMs) and propose an 11-condition cognitive evaluation framework to benchmark seven off-the-shelf LLMs in the context of authorial style emulation. We hypothesize that effective cognitive representations can help LLMs generate storytelling that better mirrors the original author. Thus, we test different cognitive representations, e.g., linguistic features, concept mappings, and profile-based information. Results show that combining conceptual and linguistic features is particularly effective in ICS, outperforming static profile-based cues in overall evaluation. Importantly, LLMs are more effective at mimicking linguistic style than narrative structure, underscoring their limits in deeper cognitive simulation. These findings provide a foundation for developing AI systems that adapt to individual ways of thinking and expression, advancing more personalized and human-aligned creative technologies.},

note = {arXiv:2510.20252 [cs.AI]},

keywords = {AI, LLM, Natural Language, Simulation},

pubstate = {published},

tppubtype = {misc}

}

@misc{kwon_evaluating_2025,

title = {Evaluating Behavioral Alignment in Conflict Dialogue: A Multi-Dimensional Comparison of LLM Agents and Humans},

author = {Deuksin Kwon and Kaleen Shrestha and Bin Han and Elena Hayoung Lee and Gale Lucas},

url = {http://arxiv.org/abs/2509.16394},

doi = {10.48550/arXiv.2509.16394},

year  = {2025},

date = {2025-09-01},

urldate = {2025-09-25},

publisher = {arXiv},

abstract = {Large Language Models (LLMs) are increasingly deployed in socially complex, interaction-driven tasks, yet their ability to mirror human behavior in emotionally and strategically complex contexts remains underexplored. This study assesses the behavioral alignment of personality-prompted LLMs in adversarial dispute resolution by simulating multi-turn conflict dialogues that incorporate negotiation. Each LLM is guided by a matched Five-Factor personality profile to control for individual variation and enhance realism. We evaluate alignment across three dimensions: linguistic style, emotional expression (e.g., anger dynamics), and strategic behavior. GPT-4.1 achieves the closest alignment with humans in linguistic style and emotional dynamics, while Claude-3.7-Sonnet best reflects strategic behavior. Nonetheless, substantial alignment gaps persist. Our findings establish a benchmark for alignment between LLMs and humans in socially complex interactions, underscoring both the promise and the limitations of personality conditioning in dialogue modeling.},

note = {arXiv:2509.16394 [cs]},

keywords = {AI, DTIC, LLM},

pubstate = {published},

tppubtype = {misc}

}

@misc{han_can_2025,

title = {Can LLMs Generate Behaviors for Embodied Virtual Agents Based on Personality Traits?},

author = {Bin Han and Deuksin Kwon and Spencer Lin and Kaleen Shrestha and Jonathan Gratch},

url = {http://arxiv.org/abs/2508.21087},

doi = {10.48550/arXiv.2508.21087},

year  = {2025},

date = {2025-08-01},

urldate = {2025-09-18},

publisher = {arXiv},

abstract = {This study proposes a framework that employs personality prompting with Large Language Models to generate verbal and nonverbal behaviors for virtual agents based on personality traits. Focusing on extraversion, we evaluated the system in two scenarios: negotiation and ice breaking, using both introverted and extroverted agents. In Experiment 1, we conducted agent to agent simulations and performed linguistic analysis and personality classification to assess whether the LLM generated language reflected the intended traits and whether the corresponding nonverbal behaviors varied by personality. In Experiment 2, we carried out a user study to evaluate whether these personality aligned behaviors were consistent with their intended traits and perceptible to human observers. Our results show that LLMs can generate verbal and nonverbal behaviors that align with personality traits, and that users are able to recognize these traits through the agents' behaviors. This work underscores the potential of LLMs in shaping personality aligned virtual agents.},

note = {arXiv:2508.21087 [cs]},

keywords = {LLM, Virtual Agents},

pubstate = {published},

tppubtype = {misc}

}

@misc{hans_quantifying_2025,

title = {Quantifying Loss Aversion in Cyber Adversaries via LLM Analysis},

author = {Soham Hans and Nikolos Gurney and Stacy Marsella and Sofia Hirschmann},

url = {http://arxiv.org/abs/2508.13240},

doi = {10.48550/arXiv.2508.13240},

year  = {2025},

date = {2025-08-01},

urldate = {2025-09-18},

publisher = {arXiv},

abstract = {Understanding and quantifying human cognitive biases from empirical data has long posed a formidable challenge, particularly in cybersecurity, where defending against unknown adversaries is paramount. Traditional cyber defense strategies have largely focused on fortification, while some approaches attempt to anticipate attacker strategies by mapping them to cognitive vulnerabilities, yet they fall short in dynamically interpreting attacks in progress. In recognition of this gap, IARPA's ReSCIND program seeks to infer, defend against, and even exploit attacker cognitive traits. In this paper, we present a novel methodology that leverages large language models (LLMs) to extract quantifiable insights into the cognitive bias of loss aversion from hacker behavior. Our data are collected from an experiment in which hackers were recruited to attack a controlled demonstration network. We process the hacker generated notes using LLMs using it to segment the various actions and correlate the actions to predefined persistence mechanisms used by hackers. By correlating the implementation of these mechanisms with various operational triggers, our analysis provides new insights into how loss aversion manifests in hacker decision-making. The results demonstrate that LLMs can effectively dissect and interpret nuanced behavioral patterns, thereby offering a transformative approach to enhancing cyber defense strategies through real-time, behavior-based analysis.},

note = {arXiv:2508.13240 [cs]},

keywords = {DTIC, LLM},

pubstate = {published},

tppubtype = {misc}

}

@misc{chaubey_face-llava_2025,

title = {Face-LLaVA: Facial Expression and Attribute Understanding through Instruction Tuning},

author = {Ashutosh Chaubey and Xulang Guan and Mohammad Soleymani},

url = {https://arxiv.org/abs/2504.07198},

doi = {10.48550/ARXIV.2504.07198},

year  = {2025},

date = {2025-04-01},

urldate = {2025-04-15},

publisher = {arXiv},

abstract = {The human face plays a central role in social communication, necessitating the use of performant computer vision tools for human-centered applications. We propose Face-LLaVA, a multimodal large language model for face-centered, in-context learning, including facial expression and attribute recognition. Additionally, Face-LLaVA is able to generate natural language descriptions that can be used for reasoning. Leveraging existing visual databases, we first developed FaceInstruct-1M, a face-centered database for instruction tuning MLLMs for face processing. We then developed a novel face-specific visual encoder powered by Face-Region Guided Cross-Attention that integrates face geometry with local visual features. We evaluated the proposed method across nine different datasets and five different face processing tasks, including facial expression recognition, action unit detection, facial attribute detection, age estimation and deepfake detection. Face-LLaVA achieves superior results compared to existing open-source MLLMs and competitive performance compared to commercial solutions. Our model output also receives a higher reasoning rating by GPT under a zero-shot setting across all the tasks. Both our dataset and model wil be released at https://face-llava.github.io to support future advancements in social AI and foundational vision-language research.},

note = {Version Number: 1},

keywords = {DTIC, LLM},

pubstate = {published},

tppubtype = {misc}

}

@misc{tak_mechanistic_2025,

title = {Mechanistic Interpretability of Emotion Inference in Large Language Models},

author = {Ala N. Tak and Amin Banayeeanzade and Anahita Bolourani and Mina Kian and Robin Jia and Jonathan Gratch},

url = {http://arxiv.org/abs/2502.05489},

doi = {10.48550/arXiv.2502.05489},

year  = {2025},

date = {2025-02-01},

urldate = {2025-02-20},

publisher = {arXiv},

abstract = {Large language models (LLMs) show promising capabilities in predicting human emotions from text. However, the mechanisms through which these models process emotional stimuli remain largely unexplored. Our study addresses this gap by investigating how autoregressive LLMs infer emotions, showing that emotion representations are functionally localized to specific regions in the model. Our evaluation includes diverse model families and sizes and is supported by robustness checks. We then show that the identified representations are psychologically plausible by drawing on cognitive appraisal theory, a well-established psychological framework positing that emotions emerge from evaluations (appraisals) of environmental stimuli. By causally intervening on construed appraisal concepts, we steer the generation and show that the outputs align with theoretical and intuitive expectations. This work highlights a novel way to causally intervene and precisely shape emotional text generation, potentially benefiting safety and alignment in sensitive affective domains.},

note = {arXiv:2502.05489 [cs]},

keywords = {DTIC, LLM},

pubstate = {published},

tppubtype = {misc}

}

@misc{brun_exploring_2025,

title = {Exploring Emotion-Sensitive LLM-Based Conversational AI},

author = {Antonin Brun and Ruying Liu and Aryan Shukla and Frances Watson and Jonathan Gratch},

url = {http://arxiv.org/abs/2502.08920},

doi = {10.48550/arXiv.2502.08920},

year  = {2025},

date = {2025-02-01},

urldate = {2025-02-20},

publisher = {arXiv},

abstract = {Conversational AI chatbots have become increasingly common within the customer service industry. Despite improvements in their emotional development, they often lack the authenticity of real customer service interactions or the competence of service providers. By comparing emotion-sensitive and emotion-insensitive LLM-based chatbots across 30 participants, we aim to explore how emotional sensitivity in chatbots influences perceived competence and overall customer satisfaction in service interactions. Additionally, we employ sentiment analysis techniques to analyze and interpret the emotional content of user inputs. We highlight that perceptions of chatbot trustworthiness and competence were higher in the case of the emotion-sensitive chatbot, even if issue resolution rates were not affected. We discuss implications of improved user satisfaction from emotion-sensitive chatbots and potential applications in support services.},

note = {arXiv:2502.08920 [cs]},

keywords = {AI, LLM},

pubstate = {published},

tppubtype = {misc}

}

@misc{walsh_fine-tuning_2025,

title = {Fine-tuning for Better Few Shot Prompting: An Empirical Comparison for Short Answer Grading},

author = {Joel Walsh and Siddarth Mamidanna and Benjamin Nye and Mark Core and Daniel Auerbach},

url = {https://arxiv.org/abs/2508.04063},

doi = {10.48550/ARXIV.2508.04063},

year  = {2025},

date = {2025-01-01},

urldate = {2025-08-19},

publisher = {arXiv},

abstract = {Research to improve Automated Short Answer Grading has recently focused on Large Language Models (LLMs) with prompt engineering and no- or few-shot prompting to achieve best results. This is in contrast to the fine-tuning approach, which has historically required large-scale compute clusters inaccessible to most users. New closed-model approaches such as OpenAI's fine-tuning service promise results with as few as 100 examples, while methods using open weights such as quantized low-rank adaptive (QLORA) can be used to fine-tune models on consumer GPUs. We evaluate both of these fine-tuning methods, measuring their interaction with few-shot prompting for automated short answer grading (ASAG) with structured (JSON) outputs. Our results show that finetuning with small amounts of data has limited utility for Llama open-weight models, but that fine-tuning methods can outperform few-shot baseline instruction-tuned LLMs for OpenAI's closed models. While our evaluation set is limited, we find some evidence that the observed benefits of finetuning may be impacted by the domain subject matter. Lastly, we observed dramatic improvement with the LLama 3.1 8B-Instruct open-weight model by seeding the initial training examples with a significant amount of cheaply generated synthetic training data.},

note = {Version Number: 1},

keywords = {DTIC, Learning Sciences, LLM, Machine Learning, UARC},

pubstate = {published},

tppubtype = {misc}

}

@misc{xue_descrip3d_2025,

title = {Descrip3D: Enhancing Large Language Model-based 3D Scene Understanding with Object-Level Text Descriptions},

author = {Jintang Xue and Ganning Zhao and Jie-En Yao and Hong-En Chen and Yue Hu and Meida Chen and Suya You and C. -C. Jay Kuo},

url = {https://arxiv.org/abs/2507.14555},

doi = {10.48550/ARXIV.2507.14555},

year  = {2025},

date = {2025-01-01},

urldate = {2025-08-19},

publisher = {arXiv},

abstract = {Understanding 3D scenes goes beyond simply recognizing objects; it requires reasoning about the spatial and semantic relationships between them. Current 3D scene-language models often struggle with this relational understanding, particularly when visual embeddings alone do not adequately convey the roles and interactions of objects. In this paper, we introduce Descrip3D, a novel and powerful framework that explicitly encodes the relationships between objects using natural language. Unlike previous methods that rely only on 2D and 3D embeddings, Descrip3D enhances each object with a textual description that captures both its intrinsic attributes and contextual relationships. These relational cues are incorporated into the model through a dual-level integration: embedding fusion and prompt-level injection. This allows for unified reasoning across various tasks such as grounding, captioning, and question answering, all without the need for task-specific heads or additional supervision. When evaluated on five benchmark datasets, including ScanRefer, Multi3DRefer, ScanQA, SQA3D, and Scan2Cap, Descrip3D consistently outperforms strong baseline models, demonstrating the effectiveness of language-guided relational representation for understanding complex indoor scenes.},

note = {Version Number: 1},

keywords = {LLM},

pubstate = {published},

tppubtype = {misc}

}

@misc{addison_c-fedrag_2024,

title = {C-FedRAG: A Confidential Federated Retrieval-Augmented Generation System},

author = {Parker Addison and Minh-Tuan H. Nguyen and Tomislav Medan and Jinali Shah and Mohammad T. Manzari and Brendan McElrone and Laksh Lalwani and Aboli More and Smita Sharma and Holger R. Roth and Isaac Yang and Chester Chen and Daguang Xu and Yan Cheng and Andrew Feng and Ziyue Xu},

url = {http://arxiv.org/abs/2412.13163},

doi = {10.48550/arXiv.2412.13163},

year  = {2024},

date = {2024-12-01},

urldate = {2025-03-20},

publisher = {arXiv},

abstract = {Organizations seeking to utilize Large Language Models (LLMs) for knowledge querying and analysis often encounter challenges in maintaining an LLM fine-tuned on targeted, up-to-date information that keeps answers relevant and grounded. Retrieval Augmented Generation (RAG) has quickly become a feasible solution for organizations looking to overcome the challenges of maintaining proprietary models and to help reduce LLM hallucinations in their query responses. However, RAG comes with its own issues regarding scaling data pipelines across tiered-access and disparate data sources. In many scenarios, it is necessary to query beyond a single data silo to provide richer and more relevant context for an LLM. Analyzing data sources within and across organizational trust boundaries is often limited by complex data-sharing policies that prohibit centralized data storage, therefore, inhibit the fast and effective setup and scaling of RAG solutions. In this paper, we introduce Confidential Computing (CC) techniques as a solution for secure Federated Retrieval Augmented Generation (FedRAG). Our proposed Confidential FedRAG system (C-FedRAG) enables secure connection and scaling of a RAG workflows across a decentralized network of data providers by ensuring context confidentiality. We also demonstrate how to implement a C-FedRAG system using the NVIDIA FLARE SDK and assess its performance using the MedRAG toolkit and MIRAGE benchmarking dataset.},

note = {arXiv:2412.13163 [cs]},

keywords = {LLM},

pubstate = {published},

tppubtype = {misc}

}