140 papers
Based on a survey of over 4,000 AI researchers, this paper reveals that contrary to public and media narratives dominated by existential threats, researchers prioritize immediate sociotechnical risks and show significant convergence with public opinion on risk assessment, suggesting a need for collaborative dialogue focused on mitigating present-day harms rather than speculating on future catastrophes.
SemFuzz is a semantics-aware fuzzing framework that leverages large language models to extract rules from RFC documents and generate targeted test cases, successfully identifying deep semantic vulnerabilities in network protocol implementations that traditional methods often miss.
This paper introduces THETA, an open-source framework that combines domain-adaptive fine-tuned embedding models with an AI Scientist Agent system to overcome the scalability and semantic limitations of traditional qualitative research, enabling rigorous, reproducible, and theoretically grounded analysis of massive social datasets.
This paper presents a user study demonstrating that human perceptions of fairness in AI systems are shaped not just by statistical parity or outcomes, but significantly by beliefs about the underlying causes of disparities, specifically how infra-marginality and data distribution differences influence judgments in medical decision-making scenarios.
Through design workshops with 15 UX designers, this paper argues that AI adoption is not merely an efficiency-driven process but a complex negotiation of values across individual, team, and organizational scales that reconfigures roles, power dynamics, and worker agency.
This paper argues that current generative AI chatbots' risk-avoidant responses to mental health crises can harm users and proposes shifting toward empowerment-oriented design principles that enable AI to act as a supportive bridge for de-escalation and connection to professional care.
This study proposes a roadmap for integrating biometric technologies into personalized Augmentative and Alternative Communication (AAC) systems by introducing concepts like the AAC biometric register, while highlighting through case studies that current AI accuracy in gesture and sign language recognition remains insufficient for practical applications and offering recommendations to bridge this gap.
This paper demonstrates that commonly used bias metrics fail to reliably capture allocational harms from large language models because they overlook the critical discrepancy between model predictions and the actual decisions made when allocating limited resources.
This study demonstrates that ChatGPT outperformed two medical students in answering challenging cardiology and vascular pathology questions, achieving a 92.10% accuracy rate on a 190-question dataset, thereby highlighting its potential as a valuable tool for medical education.
This paper presents a multilingual audit revealing that while contemporary large language models generally align with public opinion on broad social issues across Asian regions, they consistently fail to accurately represent diverse religious viewpoints—particularly those of minority groups—and often amplify negative stereotypes, a problem that persists despite lightweight prompting interventions and remains undetected by standard bias benchmarks.
This paper introduces Aletheia, an autonomous AI research agent powered by advanced reasoning models and tool use that successfully generates, verifies, and revises mathematical proofs from Olympiad problems to PhD-level research, achieving milestones such as fully AI-generated papers and the autonomous solution of open problems while proposing new frameworks for quantifying AI autonomy and transparency.
This paper introduces a framework for finetuning Large Language Models to generate age-appropriate, factually accurate educational content across six grade levels, which significantly improves grade-level alignment by 35.64 percentage points compared to standard prompting methods.
This paper proposes a unified instructional framework that integrates MOOCs, Smart Teaching, and AI into a coherent, teaching-driven pedagogy, formalizing them as a layered knowledge transformation model to maximize systemic educational potential through structured exposure, adaptive allocation, and efficiency amplification.
This paper introduces AdAEM, a novel self-extensible evaluation framework that automatically generates adaptive test questions by probing the internal value boundaries of diverse LLMs to overcome the limitations of static benchmarks and provide more informative, distinguishable insights into models' value differences and alignment dynamics.
This paper adopts a survivor-centered approach to expose how a "malicious technical ecosystem" of accessible tools enables the creation of AI-generated non-consensual intimate images, while demonstrating that current governance frameworks, such as the NIST AI 100-4 report, fail to effectively regulate this landscape due to flawed underlying assumptions.
This systematic literature review critiques the "ground truth" paradigm in machine learning as a positivistic fallacy that misinterprets human disagreement as noise, arguing instead for pluralistic annotation infrastructures that treat diverse subjective perspectives as high-fidelity signals essential for building culturally competent models.
This paper argues that while Large Language Models lack the autonomy required for genuine agency due to their absence of self-generated norms, goals, and embodied interaction, they function as transformative "linguistic automata" that, through a unique human-machine coupling, enable new "midtended" forms of intentional agency.
This paper introduces the concept of "ambiguity collapse," where large language models prematurely resolve genuinely ambiguous, value-laden terms into singular interpretations, and proposes a taxonomy of the resulting epistemic risks across process, output, and ecosystem levels alongside multi-layered mitigation strategies to preserve meaningful ambiguity.
This paper presents an algorithmic audit of TikTok revealing that while the platform technically complies with the Digital Service Act's ban on profiled advertising to minors, it effectively circumvents this protection by delivering highly personalized, often undisclosed influencer marketing content to adolescents, thereby highlighting the urgent need to expand the regulatory definition of "advertisement" to cover such commercial practices.
This paper presents a multi-source qualitative study that identifies four key themes—AI Governance and Human Authority, Human-in-the-Loop Iterative Refinement, AI System Lifecycle and Operational Constraints, and Human-AI Team Collaboration and Coordination—to address the fragmented operational guidance for structuring human roles and oversight in AI application development.