Imagen de Portada

The Dog the Cat Chased Stumped the Model: Measuring When Language Models Abandon Structure for Shortcuts

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Madhusudan, Sangmitra, Chen, Kaige, Emami, Ali
Formato:	Preprint
Publicado:	2025
Materias:	Computation and Language Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

STOP! Benchmarking Large Language Models with Sensitivity Testing on Offensive Progressions
por: Morabito, Robert, et al.
Publicado: (2024)

NYT-Connections: A Deceptively Simple Text Classification Task that Stumps System-1 Thinkers
por: Lopez, Angel Yahir Loredo, et al.
Publicado: (2024)

Which Words Matter Most in Zero-Shot Prompts?
por: Sadr, Nikta Gohari, et al.
Publicado: (2025)

Subtle Biases Need Subtler Measures: Dual Metrics for Evaluating Representative and Affinity Bias in Large Language Models
por: Kumar, Abhishek, et al.
Publicado: (2024)

Fine-Tuned LLMs are "Time Capsules" for Tracking Societal Bias Through Books
por: Madhusudan, Sangmitra, et al.
Publicado: (2025)

MirrorStories: Reflecting Diversity through Personalized Narrative Generation with Large Language Models
por: Yunusov, Sarfaroz, et al.
Publicado: (2024)

Metamorphic Testing for Fairness Evaluation in Large Language Models: Identifying Intersectional Bias in LLaMA and GPT
por: Reddy, Harishwar, et al.
Publicado: (2025)

Think Twice: Measuring the Efficiency of Eliminating Prediction Shortcuts of Question Answering Models
por: Mikula, Lukáš, et al.
Publicado: (2023)

Break the Chain: Large Language Models Can be Shortcut Reasoners
por: Ding, Mengru, et al.
Publicado: (2024)

Common to Whom? Regional Cultural Commonsense and LLM Bias in India
por: Madhusudan, Sangmitra, et al.
Publicado: (2026)

TALE: A Tool-Augmented Framework for Reference-Free Evaluation of Large Language Models
por: Badshah, Sher, et al.
Publicado: (2025)

Mitigating Shortcut Reasoning in Language Models: A Gradient-Aware Training Approach
por: Cao, Hongyu, et al.
Publicado: (2026)

Learning Shortcuts: On the Misleading Promise of NLU in Language Models
por: Bihani, Geetanjali, et al.
Publicado: (2024)

Confidence Under the Hood: An Investigation into the Confidence-Probability Alignment in Large Language Models
por: Kumar, Abhishek, et al.
Publicado: (2024)

Trace-of-Thought Prompting: Investigating Prompt-Based Knowledge Distillation Through Question Decomposition
por: McDonald, Tyler, et al.
Publicado: (2025)

Knowing the Facts but Choosing the Shortcut: Understanding How Large Language Models Compare Entities
por: Lehmann, Hans Hergen, et al.
Publicado: (2025)

Spurious Correlations and Beyond: Understanding and Mitigating Shortcut Learning in SDOH Extraction with Large Language Models
por: Sakib, Fardin Ahsan, et al.
Publicado: (2025)

Embedding-Informed Adaptive Retrieval-Augmented Generation of Large Language Models
por: Huang, Chengkai, et al.
Publicado: (2024)

Chimera: Diagnosing Shortcut Learning in Visual-Language Understanding
por: Chi, Ziheng, et al.
Publicado: (2025)

The Confidence Shortcut: A Reasoning Failure Mode of Masked Diffusion Models
por: Kim, Dueun, et al.
Publicado: (2026)

Efficient Fairness Testing in Large Language Models: Prioritizing Metamorphic Relations for Bias Detection
por: Giramata, Suavis, et al.
Publicado: (2025)

The Reliability Paradox: Exploring How Shortcut Learning Undermines Language Model Calibration
por: Bihani, Geetanjali, et al.
Publicado: (2024)

Deceptive Semantic Shortcuts on Reasoning Chains: How Far Can Models Go without Hallucination?
por: Li, Bangzheng, et al.
Publicado: (2023)

When Do Tools and Planning Help Large Language Models Think? A Cost- and Latency-Aware Benchmark
por: Ghoshal, Subha, et al.
Publicado: (2026)

The Accuracy Paradox in RLHF: When Better Reward Models Don't Yield Better Language Models
por: Chen, Yanjun, et al.
Publicado: (2024)

When Large Language Models contradict humans? Large Language Models' Sycophantic Behaviour
por: Ranaldi, Leonardo, et al.
Publicado: (2023)

SkipCat: Rank-Maximized Low-Rank Compression of Large Language Models via Shared Projection and Block Skipping
por: Lu, Yu-Chen, et al.
Publicado: (2025)

When Quantization Affects Confidence of Large Language Models?
por: Proskurina, Irina, et al.
Publicado: (2024)

CURE: Controlled Unlearning for Robust Embeddings -- Mitigating Conceptual Shortcuts in Pre-Trained Language Models
por: Kocak, Aysenur, et al.
Publicado: (2025)

When Chain of Thought is Necessary, Language Models Struggle to Evade Monitors
por: Emmons, Scott, et al.
Publicado: (2025)

Rectifying Demonstration Shortcut in In-Context Learning
por: Jang, Joonwon, et al.
Publicado: (2024)

Cognitive Decision Routing in Large Language Models: When to Think Fast, When to Think Slow
por: Du, Y., et al.
Publicado: (2025)

Chasing Random: Instruction Selection Strategies Fail to Generalize
por: Diddee, Harshita, et al.
Publicado: (2024)

Measuring Form and Function in Language Models
por: Martínez, Héctor Javier Vázquez, et al.
Publicado: (2026)

Large Language Models Often Know When They Are Being Evaluated
por: Needham, Joe, et al.
Publicado: (2025)

When Language Overrules: Revealing Text Dominance in Multimodal Large Language Models
por: Wu, Huyu, et al.
Publicado: (2025)

Adaptive Large Language Models By Layerwise Attention Shortcuts
por: Verma, Prateek, et al.
Publicado: (2024)

CFDLLMBench: A Benchmark Suite for Evaluating Large Language Models in Computational Fluid Dynamics
por: Somasekharan, Nithin, et al.
Publicado: (2025)

Shortcut Learning in In-Context Learning: A Survey
por: Song, Rui, et al.
Publicado: (2024)

On the Shortcut Learning in Multilingual Neural Machine Translation
por: Wang, Wenxuan, et al.
Publicado: (2024)