Image de couverture de livre

"They parted illusions -- they parted disclaim marinade": Misalignment as structural fidelity in LLMs

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteur principal:	Costa, Mariana Lins
Format:	Preprint
Publié:	2025
Sujets:	Artificial Intelligence Computation and Language Computers and Society
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Misaligned by Reward: Socially Undesirable Preferences in LLMs
par: Ghazaryan, Gayane, et autres
Publié: (2026)

The Ghost in the Grammar: Methodological Anthropomorphism in AI Safety Evaluations
par: Costa, Mariana Lins
Publié: (2026)

SteLLA: A Structured Grading System Using LLMs with RAG
par: Qiu, Hefei, et autres
Publié: (2025)

Assessing the Performance of Human-Capable LLMs -- Are LLMs Coming for Your Job?
par: Mavi, John, et autres
Publié: (2024)

The simulation of judgment in LLMs
par: Loru, Edoardo, et autres
Publié: (2025)

Measuring Teaching with LLMs
par: Hardy, Michael
Publié: (2025)

The Political Preferences of LLMs
par: Rozado, David
Publié: (2024)

Why are all LLMs Obsessed with Japanese Culture? On the Hidden Cultural and Regional Biases of LLMs
par: de Landa, Joseba Fernandez, et autres
Publié: (2026)

Moral Mazes in the Era of LLMs
par: Nguyen, Dang, et autres
Publié: (2026)

Are LLMs (Really) Ideological? An IRT-based Analysis and Alignment Tool for Perceived Socio-Economic Bias in LLMs
par: Wachter, Jasmin, et autres
Publié: (2025)

Interpretability Framework for LLMs in Undergraduate Calculus
par: Dakshit, Sagnik, et autres
Publié: (2025)

MEDEQUALQA: Evaluating Biases in LLMs with Counterfactual Reasoning
par: Ghosh, Rajarshi, et autres
Publié: (2025)

ELEPHANT: Measuring and understanding social sycophancy in LLMs
par: Cheng, Myra, et autres
Publié: (2025)

On the Credibility of Evaluating LLMs using Survey Questions
par: Libovický, Jindřich
Publié: (2026)

Verbalizing LLMs' assumptions to explain and control sycophancy
par: Cheng, Myra, et autres
Publié: (2026)

EtiCor++: Towards Understanding Etiquettical Bias in LLMs
par: Dwivedi, Ashutosh, et autres
Publié: (2025)

EulerESG: Automating ESG Disclosure Analysis with LLMs
par: Ding, Yi, et autres
Publié: (2025)

Large Language Models (LLMs) as Agents for Augmented Democracy
par: Gudiño-Rosero, Jairo, et autres
Publié: (2024)

Automated Assessment of Students' Code Comprehension using LLMs
par: Oli, Priti, et autres
Publié: (2023)

Evaluating the Capabilities of LLMs for Supporting Anticipatory Impact Assessment
par: Allaham, Mowafak, et autres
Publié: (2024)

Explore the Potential of LLMs in Misinformation Detection: An Empirical Study
par: Chen, Mengyang, et autres
Publié: (2023)

Perceived Political Bias in LLMs Reduces Persuasive Abilities
par: DiGiuseppe, Matthew, et autres
Publié: (2026)

Evaluating Cultural Awareness of LLMs for Yoruba, Malayalam, and English
par: Dawson, Fiifi, et autres
Publié: (2024)

Towards Measuring and Modeling "Culture" in LLMs: A Survey
par: Adilazuarda, Muhammad Farid, et autres
Publié: (2024)

Fluent but Foreign: Even Regional LLMs Lack Cultural Alignment
par: Agarwal, Dhruv, et autres
Publié: (2025)

HugAgent: Benchmarking LLMs for Simulation of Individualized Human Reasoning
par: Li, Chance Jiajie, et autres
Publié: (2025)

What Is The Political Content in LLMs' Pre- and Post-Training Data?
par: Ceron, Tanise, et autres
Publié: (2025)

Mitigating Gender Bias via Fostering Exploratory Thinking in LLMs
par: Wei, Kangda, et autres
Publié: (2025)

HumT DumT: Measuring and controlling human-like language in LLMs
par: Cheng, Myra, et autres
Publié: (2025)

The Homogenization Problem in LLMs: Towards Meaningful Diversity in AI Safety
par: Rios-Sialer, Ian
Publié: (2026)

Empathy and the Right to Be an Exception: What LLMs Can and Cannot Do
par: Kidder, William, et autres
Publié: (2024)

Widespread Gender and Pronoun Bias in Moral Judgments Across LLMs
par: Fernandes, Gustavo Lúcius, et autres
Publié: (2026)

Are Rationales Necessary and Sufficient? Tuning LLMs for Explainable Misinformation Detection
par: Wang, Bing, et autres
Publié: (2026)

Break the Checkbox: Challenging Closed-Style Evaluations of Cultural Alignment in LLMs
par: Kabir, Mohsinul, et autres
Publié: (2025)

Are LLMs Court-Ready? Evaluating Frontier Models on Indian Legal Reasoning
par: Juvekar, Kush, et autres
Publié: (2025)

LocalBench: Benchmarking LLMs on County-Level Local Knowledge and Reasoning
par: Gao, Zihan, et autres
Publié: (2025)

AdAEM: An Adaptively and Automated Extensible Measurement of LLMs' Value Difference
par: Yao, Jing, et autres
Publié: (2025)

H-Neurons: On the Existence, Impact, and Origin of Hallucination-Associated Neurons in LLMs
par: Gao, Cheng, et autres
Publié: (2025)

Safe in the Future, Dangerous in the Past: Dissecting Temporal and Linguistic Vulnerabilities in LLMs
par: Said, Muhammad Abdullahi, et autres
Publié: (2025)

Show, Don't Tell: Uncovering Implicit Character Portrayal using LLMs
par: Jaipersaud, Brandon, et autres
Publié: (2024)