Image de couverture de livre

Safety Drift After Fine-Tuning: Evidence from High-Stakes Domains

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Khan, Emaan Bilal, Winecoff, Amy, Bogen, Miranda, Hadfield-Menell, Dylan
Format:	Preprint
Publié:	2026
Sujets:	Computers and Society Software Engineering
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Assessing a Safety Case: Bottom-up Guidance for Claims and Evidence Evaluation
par: Schnelle, Scott, et autres
Publié: (2025)

Pitfalls of Evidence-Based AI Policy
par: Casper, Stephen, et autres
Publié: (2025)

Randomness, Not Representation: The Unreliability of Evaluating Cultural Alignment in LLMs
par: Khan, Ariba, et autres
Publié: (2025)

Measuring What Matters: A Framework for Evaluating Safety Risks in Real-World LLM Applications
par: Goh, Jia Yi, et autres
Publié: (2025)

Improving governance outcomes through AI documentation: Bridging theory and practice
par: Winecoff, Amy A., et autres
Publié: (2024)

LLM Contribution Summarization in Software Projects
par: Ferrao, Rafael Corsi, et autres
Publié: (2025)

Narrowing the Gap: Supervised Fine-Tuning of Open-Source LLMs as a Viable Alternative to Proprietary Models for Pedagogical Tools
par: Solano, Lorenzo Lee, et autres
Publié: (2025)

From Generation to Adaptation: Comparing AI-Assisted Strategies in High School Programming Education
par: Hu, Tong, et autres
Publié: (2025)

Zero-Shot Cross-Domain Code Search without Fine-Tuning
par: Liang, Keyu, et autres
Publié: (2025)

PRISM: A Design Framework for Open-Source Foundation Model Safety
par: Neumann, Terrence, et autres
Publié: (2024)

Insights from Publishing Open Data in Industry-Academia Collaboration
par: Strandberg, Per Erik, et autres
Publié: (2025)

Early Results from Teaching Modelling for Software Comprehension in New-Hire Onboarding
par: Kumar, Mrityunjay, et autres
Publié: (2025)

ChatGPT and Its Educational Impact: Insights from a Software Development Competition
par: Hwang, Sunhee, et autres
Publié: (2024)

From Pre-labeling to Production: Engineering Lessons from a Machine Learning Pipeline in the Public Sector
par: Ferreira, Ronivaldo, et autres
Publié: (2025)

Assessing Simulation Knowledge and Proficiency Among Undergraduate Computing Students in Brazil: Insights and Results from a Survey Research
par: Rodrigues, Fernando Brito, et autres
Publié: (2025)

CALMA: A Process for Deriving Context-aligned Axes for Language Model Alignment
par: Soni, Prajna, et autres
Publié: (2025)

LLM Use, Cheating, and Academic Integrity in Software Engineering Education
par: Santos, Ronnie de Souza, et autres
Publié: (2026)

Understanding: reframing automation and assurance
par: Bloomfield, Robin
Publié: (2026)

Bringing AI into the Classroom: A Structured Approach for Integrating AI into Software Engineering Education
par: Groher, Iris, et autres
Publié: (2026)

The Need for a Green ICT Reference Framework
par: Aiello, Marco, et autres
Publié: (2026)

Operationalizing Ethics for AI Agents: How Developers Encode Values into Repository Context Files
par: Treude, Christoph, et autres
Publié: (2026)

Privacy-by-Design Adaptive Group Assignment for Digital Lifestyle Coaching at Scale
par: Mani, Nariman, et autres
Publié: (2026)

Putting green software principles into practice
par: Uther, James
Publié: (2026)

Towards a Goal-Centric Assessment of Requirements Engineering Methods for Privacy by Design
par: Kosenkov, Oleksandr, et autres
Publié: (2026)

They Call Her 'Miss' and Him 'Professor': Lived Experiences of Women Teaching Support Staff in IT/SE Education
par: Malhotra, Vasudha, et autres
Publié: (2026)

"Write in English, Nobody Understands Your Language Here": A Study of Non-English Trends in Open-Source Repositories
par: Bhuiyan, Masudul Hasan Masud, et autres
Publié: (2026)

Teaching Empathy in Software Engineering Education in the Age of Artificial Intelligence
par: Santos, Ronnie de Souza, et autres
Publié: (2026)

Integrating Mental Health, Well-Being, and Sustainability into Software Engineering Education
par: Graßl, Isabella, et autres
Publié: (2026)

Learning to Program Alongside AI: Critical Thinking, AI Ethics, and Gendered Patterns of German Secondary School Students
par: Graßl, Isabella
Publié: (2026)

Generative AI Adoption in an Energy Company: Exploring Challenges and Use Cases
par: Sami, Malik Abdul, et autres
Publié: (2026)

An Empirical Analysis of Community and Coding Patterns in OSS4SG vs. Conventional OSS
par: Ouf, Mohamed, et autres
Publié: (2026)

Beyond the Binary: Motivations, Challenges, and Strategies of Transgender and Non-binary Software Engineering Students
par: Graßl, Isabella
Publié: (2026)

Design and Deployment of a Course-Aware AI Tutor in an Introductory Programming Course
par: Groher, Iris, et autres
Publié: (2026)

What Should Frontier AI Developers Disclose About Internal Deployments?
par: Charnock, Jacob, et autres
Publié: (2026)

Towards More Empathic Programming Environments: An Experimental Empathic AI-Enhanced IDE
par: Go, Justin Rainier, et autres
Publié: (2026)

The First Issue Matters: Linking Task-Level Characteristics to Long-Term Newcomer Retention in OSS
par: Hao, Yichen, et autres
Publié: (2026)

Towards Viewpoint-centric Artifact-based Regulatory Requirements Engineering for Compliance by Design
par: Kosenkov, Oleksandr
Publié: (2026)

Social Proof is in the Pudding: The (Non)-Impact of Social Proof on Software Downloads
par: Shen, Lucas, et autres
Publié: (2026)

Governance-Aware Software Architecture for Multi-Stakeholder Platforms
par: Nwankwo, Michael, et autres
Publié: (2026)

A Systematic AI Adoption Framework for Higher Education: From Student GenAI Usage to Institutional Integration
par: Neumann, Michael, et autres
Publié: (2026)