Computational Social Science (CSS) is an emerging field driven by the unprecedented availability of human-generated content for researchers. This field, however, presents a unique set of challenges due to the nature of the theories and datasets it explores, including highly subjective tasks and complex, unstructured textual corpora. Among these challenges, one of the less well-studied topics is the pervasive presence of noise. This thesis aims to address this gap in the literature by presenting a series of interconnected case studies that examine different manifestations of noise in CSS. These include character-level errors following the OCR processing of historical records, archaic language, inconsistencies in annotations for subjective and ambiguous tasks, and even noise and biases introduced by large language models during content generation. This thesis challenges the conventional notion that noise in CSS is inherently harmful or useless. Rather, it argues that certain forms of noise can encode meaningful information that is invaluable for advancing CSS research, such as the unique communication styles of individuals or the culture-dependent nature of datasets and tasks. Further, this thesis highlights the importance of nuance in dealing with noise and the considerations CSS researchers must address when encountering it, demonstrating that different types of noise require distinct strategies.

We present RepoST, a scalable method to construct environments that provide execution feedback for repository-level code generation for both training and evaluation. Unlike existing works that aim to build entire repositories for execution, which is challenging for both human and LLMs, we provide execution feedback with sandbox testing, which isolates a given target function and its dependencies to a separate script for testing. Sandbox testing reduces the complexity of external dependencies and enables constructing environments at a large scale. We use our method to construct RepoST-Train, a large-scale train set with 7,415 functions from 832 repositories. Training with the execution feedback provided by RepoST-Train leads to a performance gain of 5.5% Pass@1 on HumanEval and 3.5% Pass@1 on RepoEval. We also build an evaluation dataset, RepoST-Eval, and benchmark 12 code generation models.

The impact of random seeds in fine-tuning large language models (LLMs) has been largely overlooked despite its potential influence on model performance.In this study, we systematically evaluate the effects of random seeds on LLMs using the GLUE and SuperGLUE benchmarks. We analyze the macro-level impact through traditional metrics like accuracy and F1, calculating their mean and variance to quantify performance fluctuations. To capture the micro-level effects, we introduce a novel metric, consistency, measuring the stability of individual predictions across runs. Our experiments reveal significant variance at both macro and micro levels, underscoring the need for careful consideration of random seeds in fine-tuning and evaluation.

Large Language Models (LLMs) have demonstrated strong generalizable reasoning and planning capabilities. However, their efficacies in spatial path planning and obstacle-free trajectory generation remain underexplored. Leveraging LLMs for navigation holds significant potential, given LLMs' ability to handle unseen scenarios, support user-agent interactions, and provide global control across complex systems, making them well-suited for agentic planning and humanoid motion generation. As one of the first studies in this domain, we explore the zero-shot navigation and path generation capabilities of LLMs by constructing a dataset and proposing an evaluation protocol. Specifically, we represent paths using anchor points connected by straight lines, enabling movement in various directions. This approach offers greater flexibility and practicality compared to previous methods while remaining simple and intuitive for LLMs. We demonstrate that, when tasks are well-structured in this manner, modern LLMs exhibit substantial planning proficiency in avoiding obstacles while autonomously refining navigation with the generated motion to reach the target. Further, this spatial reasoning ability of a single LLM motion agent interacting in a static environment can be seamlessly generalized in multi-motion agents coordination in dynamic environments. Unlike traditional approaches that rely on single-step planning or local policies, our training-free LLM-based method enables global, dynamic, closed-loop planning, and autonomously resolving collision issues.

With the rise of large language models (LLMs), AI agents as autonomous decision-makers present significant opportunities and challenges for human-AI cooperation. While many studies have explored human cooperation with AI as tools, the role of LLM-augmented autonomous agents in competitive-cooperative interactions remains under-examined. This study investigates human cooperative behavior by engaging 30 participants who interacted with LLM agents exhibiting different characteristics (purported human, purported rule-based AI agent, and LLM agent) in repeated Prisoner's Dilemma games. Findings show significant differences in cooperative behavior based on the agents' purported characteristics and the interaction effect of participants' genders and purported characteristics. We also analyzed human response patterns, including game completion time, proactive favorable behavior, and acceptance of repair efforts. These insights offer a new perspective on human interactions with LLM agents in competitive cooperation contexts, such as virtual avatars or future physical entities. The study underscores the importance of understanding human biases toward AI agents and how observed behaviors can influence future human-AI cooperation dynamics.

In robot task planning, large language models (LLMs) have shown significant promise in generating complex and long-horizon action sequences. However, it is observed that LLMs often produce responses that sound plausible but are not accurate. To address these problems, existing methods typically employ predefined error sets or external knowledge sources, requiring human efforts and computation resources. Recently, self-correction approaches have emerged, where LLM generates and refines plans, identifying errors by itself. Despite their effectiveness, they are more prone to failures in correction due to insufficient reasoning. In this paper, we introduce InversePrompt, a novel self-corrective task planning approach that leverages inverse prompting to enhance interpretability. Our method incorporates reasoning steps to provide clear, interpretable feedback. It generates inverse actions corresponding to the initially generated actions and verifies whether these inverse actions can restore the system to its original state, explicitly validating the logical coherence of the generated plans. The results on benchmark datasets show an average 16.3% higher success rate over existing LLM-based task planning methods. Our approach offers clearer justifications for feedback in real-world environments, resulting in more successful task completion than existing self-correction approaches across various scenarios.

In recent years, large language models (LLMs), empowered by massive text corpora and deep learning techniques, have demonstrated breakthrough advancements in cross-domain knowledge transfer and human-machine dialogue interactions [1]. Within the healthcare domain, LLMs are increasingly deployed across nine core application scenarios, including intelligent diagnosis, personalized treatment, and drug discovery, garnering significant attention from both academia and industry [2, 3]. A particularly important area of focus is the development and evaluation of Chinese medical LLMs, which face unique challenges due to the specialized nature of medical knowledge and the high-stakes implications of clinical decision-making. Hence, ensuring the reliability and safety of these models has become critical, necessitating rigorous evaluation frameworks [4]. Current research on medical LLMs evaluation exhibits two predominant trends. On one hand, general-domain benchmarks (e.g., HELM [5], MMLU [6]) assess foundational model capabilities through medical knowledge tests. On the other hand, specialized medical evaluation systems (e.g., MedQA [7], C-Eval-Medical [8]) emphasize clinical reasoning and ethical compliance. Notably, the MedBench framework [9], jointly developed by institutions including Shanghai AI Laboratory, has emerged as the most influential benchmark for Chinese medical LLMs. By establishing a standardized evaluation system spanning five dimensions--medical language comprehension, complex reasoning, and safety ethics--it has attracted participation from hundreds of research teams.

Fact-checking plays a crucial role in combating misinformation. Existing methods using large language models (LLMs) for claim decomposition face two key limitations: (1) insufficient decomposition, introducing unnecessary complexity to the verification process, and (2) ambiguity of mentions, leading to incorrect verification results. To address these challenges, we suggest introducing a claim graph consisting of triplets to address the insufficient decomposition problem and reduce mention ambiguity through graph structure. Based on this core idea, we propose a graph-based framework, GraphFC, for fact-checking. The framework features three key components: graph construction, which builds both claim and evidence graphs; graph-guided planning, which prioritizes the triplet verification order; and graph-guided checking, which verifies the triples one by one between claim and evidence graphs. Extensive experiments show that GraphFC enables fine-grained decomposition while resolving referential ambiguities through relational constraints, achieving state-of-the-art performance across three datasets.

Trust plays a fundamental role in shaping the willingness of users to engage and collaborate with artificial intelligence (AI) systems. Yet, measuring user trust remains challenging due to its complex and dynamic nature. While traditional survey methods provide trust levels for long conversations, they fail to capture its dynamic evolution during ongoing interactions. Here, we present VizTrust, which addresses this challenge by introducing a real-time visual analytics tool that leverages a multi-agent collaboration system to capture and analyze user trust dynamics in human-agent communication. Built on established human-computer trust scales-competence, integrity, benevolence, and predictability-, VizTrust enables stakeholders to observe trust formation as it happens, identify patterns in trust development, and pinpoint specific interaction elements that influence trust. Our tool offers actionable insights into human-agent trust formation and evolution in real time through a dashboard, supporting the design of adaptive conversational agents that responds effectively to user trust signals.

Zero-shot human-AI coordination is the training of an ego-agent to coordinate with humans without using human data. Most studies on zero-shot human-AI coordination have focused on enhancing the ego-agent's coordination ability in a given environment without considering the issue of generalization to unseen environments. Real-world applications of zero-shot human-AI coordination should consider unpredictable environmental changes and the varying coordination ability of co-players depending on the environment. Previously, the multi-agent UED (Unsupervised Environment Design) approach has investigated these challenges by jointly considering environmental changes and co-player policy in competitive two-player AI-AI scenarios. In this paper, our study extends the multi-agent UED approach to a zero-shot human-AI coordination. We propose a utility function and co-player sampling for a zero-shot human-AI coordination setting that helps train the ego-agent to coordinate with humans more effectively than the previous multi-agent UED approach. The zero-shot human-AI coordination performance was evaluated in the Overcooked-AI environment, using human proxy agents and real humans. Our method outperforms other baseline models and achieves a high human-AI coordination performance in unseen environments.