Imagen de Portada

Measuring Evaluation-Context Divergence in Open-Weight LLMs: A Paired-Prompt Protocol with Pilot Evidence of Alignment-Pipeline-Specific Heterogeneity

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Burnat, Florian A. D., Davidson, Brittany I.
Formato:	Preprint
Publicado:	2026
Materias:	Computation and Language Artificial Intelligence Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

The Accountability Paradox: How Platform API Restrictions Undermine AI Transparency Mandates
por: Burnat, Florian A. D., et al.
Publicado: (2025)

Evaluating LLMs' Divergent Thinking Capabilities for Scientific Idea Generation with Minimal Context
por: Ruan, Kai, et al.
Publicado: (2024)

Bring Your Own Prompts: Use-Case-Specific Bias and Fairness Evaluation for LLMs
por: Bouchard, Dylan
Publicado: (2024)

Regulatory gray areas of LLM Terms
por: Davidson, Brittany I., et al.
Publicado: (2026)

Narrative Context Protocol: An Open-Source Storytelling Framework for Generative AI
por: Gerba, Hank
Publicado: (2025)

Context Misleads LLMs: The Role of Context Filtering in Maintaining Safe Alignment of LLMs
por: Kim, Jinhwa, et al.
Publicado: (2025)

Qworld: Question-Specific Evaluation Criteria for LLMs
por: Gao, Shanghua, et al.
Publicado: (2026)

Language Statistics and False Belief Reasoning: Evidence from 41 Open-Weight LMs
por: Trott, Sean, et al.
Publicado: (2026)

Evaluating the Sensitivity of LLMs to Prior Context
por: Hankache, Robert, et al.
Publicado: (2025)

Tamper-Resistant Safeguards for Open-Weight LLMs
por: Tamirisa, Rishub, et al.
Publicado: (2024)

DETAIL Matters: Measuring the Impact of Prompt Specificity on Reasoning in Large Language Models
por: Kim, Olivia
Publicado: (2025)

Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing
por: Xu, Zhangchen, et al.
Publicado: (2024)

Humans and LLMs Diverge on Probabilistic Inferences
por: Kamath, Gaurav, et al.
Publicado: (2026)

Adapting LLMs to Time Series Forecasting via Temporal Heterogeneity Modeling and Semantic Alignment
por: Sun, Yanru, et al.
Publicado: (2025)

Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges
por: Thakur, Aman Singh, et al.
Publicado: (2024)

Ethical Reasoning and Moral Value Alignment of LLMs Depend on the Language we Prompt them in
por: Agarwal, Utkarsh, et al.
Publicado: (2024)

Evaluating the Evaluator: Measuring LLMs' Adherence to Task Evaluation Instructions
por: Murugadoss, Bhuvanashree, et al.
Publicado: (2024)

SDA: Steering-Driven Distribution Alignment for Open LLMs without Fine-Tuning
por: Xia, Wei, et al.
Publicado: (2025)

On Evaluating LLM Alignment by Evaluating LLMs as Judges
por: Liu, Yixin, et al.
Publicado: (2025)

CompactQE: Interpretable Translation Quality Estimation via Small Open-Weight LLMs
por: Guttmann, Kamil, et al.
Publicado: (2026)

Pair-In, Pair-Out: Latent Multi-Token Prediction for Efficient LLMs
por: Tan, Wenhui, et al.
Publicado: (2026)

Systematic Evaluation of Long-Context LLMs on Financial Concepts
por: Gupta, Lavanya, et al.
Publicado: (2024)

ALI-Agent: Assessing LLMs' Alignment with Human Values via Agent-based Evaluation
por: Zheng, Jingnan, et al.
Publicado: (2024)

How Far Can In-Context Alignment Go? Exploring the State of In-Context Alignment
por: Huang, Heyan, et al.
Publicado: (2024)

TN-Eval: Rubric and Evaluation Protocols for Measuring the Quality of Behavioral Therapy Notes
por: Shah, Raj Sanjay, et al.
Publicado: (2025)

Haystack Engineering: Context Engineering for Heterogeneous and Agentic Long-Context Evaluation
por: Li, Mufei, et al.
Publicado: (2025)

Beyond Textual Context: Structural Graph Encoding with Adaptive Space Alignment to alleviate the hallucination of LLMs
por: Zhang, Yifang, et al.
Publicado: (2025)

Hallucination Detection in LLMs with Topological Divergence on Attention Graphs
por: Bazarova, Alexandra, et al.
Publicado: (2025)

Unveiling Divergent Inductive Biases of LLMs on Temporal Data
por: Kishore, Sindhu, et al.
Publicado: (2024)

Alignment Tuning for Large Language Models: A Data-Centric Lens on Alignment Data Pipelines
por: Song, Hwanjun
Publicado: (2026)

LLM-Human Pipeline for Cultural Context Grounding of Conversations
por: Pujari, Rajkumar, et al.
Publicado: (2024)

Pipelined Decoder for Efficient Context-Aware Text Generation
por: Huang, Zixian, et al.
Publicado: (2025)

Self-Alignment for Factuality: Mitigating Hallucinations in LLMs via Self-Evaluation
por: Zhang, Xiaoying, et al.
Publicado: (2024)

From Guidelines to Guarantees: A Graph-Based Evaluation Harness for Domain-Specific Evaluation of LLMs
por: Lundin, Jessica M., et al.
Publicado: (2025)

Diverge to Induce Prompting: Multi-Rationale Induction for Zero-Shot Reasoning
por: Chen, Po-Chun, et al.
Publicado: (2026)

StealthGraph: Exposing Domain-Specific Risks in LLMs through Knowledge-Graph-Guided Harmful Prompt Generation
por: Zheng, Huawei, et al.
Publicado: (2026)

OpenSanctions Pairs: Large-Scale Entity Matching with LLMs
por: Smith, Chandler, et al.
Publicado: (2026)

Evidence-backed Fact Checking using RAG and Few-Shot In-Context Learning with LLMs
por: Singhal, Ronit, et al.
Publicado: (2024)

Localizing AI: Evaluating Open-Weight Language Models for Languages of Baltic States
por: Kapočiūtė-Dzikienė, Jurgita, et al.
Publicado: (2025)

Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference
por: Chiang, Wei-Lin, et al.
Publicado: (2024)