Cover Image

Multimodal LLMs Do Not Compose Skills Optimally Across Modalities

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Ontalvilla, Paula, Ormazabal, Aitor, Azkune, Gorka
Format:	Preprint
Published:	2025
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Improving the Efficiency of Visually Augmented Language Models
by: Ontalvilla, Paula, et al.
Published: (2024)

BertaQA: How Much Do Language Models Know About Local Culture?
by: Etxaniz, Julen, et al.
Published: (2024)

When to Retrieve: Teaching LLMs to Utilize Information Retrieval Effectively
by: Labruna, Tiziano, et al.
Published: (2024)

Grounding Spatial Relations in Text-Only Language Models
by: Azkune, Gorka, et al.
Published: (2024)

Multimodal Large Language Models for Low-Resource Languages: A Case Study for Basque
by: Arana, Lukas, et al.
Published: (2025)

Vision-Language Models Struggle to Align Entities across Modalities
by: Alonso, Iñigo, et al.
Published: (2025)

Adding simple structure at inference improves Vision-Language Compositionality
by: Miranda, Imanol, et al.
Published: (2025)

BiVLC: Extending Vision-Language Compositionality Evaluation with Text-to-Image Retrieval
by: Miranda, Imanol, et al.
Published: (2024)

Revisiting Compositionality in Dual-Encoder Vision-Language Models: The Role of Inference
by: Miranda, Imanol, et al.
Published: (2026)

Conditioning LLMs to Generate Code-Switched Text
by: Heredia, Maite, et al.
Published: (2025)

Do Composed Image Retrieval Benchmarks Require Multimodal Composition?
by: Attimonelli, Matteo, et al.
Published: (2026)

MCA: Modality Composition Awareness for Robust Composed Multimodal Retrieval
by: Wu, Qiyu, et al.
Published: (2025)

Latxa: An Open Language Model and Evaluation Suite for Basque
by: Etxaniz, Julen, et al.
Published: (2024)

Can Language Models Compose Skills In-Context?
by: Liu, Zidong, et al.
Published: (2025)

Do Methods to Jailbreak and Defend LLMs Generalize Across Languages?
by: Atil, Berk, et al.
Published: (2025)

How Do Language Models Compose Functions?
by: Khandelwal, Apoorv, et al.
Published: (2025)

Red Teaming Multimodal Language Models: Evaluating Harm Across Prompt Modalities and Models
by: Van Doren, Madison, et al.
Published: (2025)

Do AI Models Perform Human-like Abstract Reasoning Across Modalities?
by: Beger, Claas, et al.
Published: (2025)

"Yeah Right!" -- Do LLMs Exhibit Multimodal Feature Transfer?
by: Reichman, Benjamin, et al.
Published: (2025)

From $f(x)$ and $g(x)$ to $f(g(x))$: LLMs Learn New Skills in RL by Composing Old Ones
by: Yuan, Lifan, et al.
Published: (2025)

Do LLMs and VLMs Share Neurons for Inference? Evidence and Mechanisms of Cross-Modal Transfer
by: Cui, Chenhang, et al.
Published: (2026)

Improving Explicit Spatial Relationships in Text-to-Image Generation through an Automatically Derived Dataset
by: Salaberria, Ander, et al.
Published: (2024)

Compose and Fuse: Revisiting the Foundational Bottlenecks in Multimodal Reasoning
by: Wang, Yucheng, et al.
Published: (2025)

Composing or Not Composing? Towards Distributional Construction Grammars
by: Blache, Philippe, et al.
Published: (2024)

Do LLMs Understand Wine Descriptors Across Cultures? A Benchmark for Cultural Adaptations of Wine Reviews
by: Zou, Chenye, et al.
Published: (2025)

Prompt Decorators: A Declarative and Composable Syntax for Reasoning, Formatting, and Control in LLMs
by: Heris, Mostapha Kalami
Published: (2025)

Skill-LLM: Repurposing General-Purpose LLMs for Skill Extraction
by: Herandi, Amirhossein, et al.
Published: (2024)

How Do Vision-Language Models Process Conflicting Information Across Modalities?
by: Hua, Tianze, et al.
Published: (2025)

Encyclo-K: Evaluating LLMs with Dynamically Composed Knowledge Statements
by: Liang, Yiming, et al.
Published: (2025)

Identifying High-Confidence Social Biases in LLMs for Trustworthy Conversational Tutoring Agents
by: Alvarez, Aitor Arronte, et al.
Published: (2026)

Leveraging LLMs For Turkish Skill Extraction
by: İltüzer, Ezgi Arslan, et al.
Published: (2026)

Language on Demand, Knowledge at Core: Composing LLMs with Encoder-Decoder Translation Models for Extensible Multilinguality
by: Bu, Mengyu, et al.
Published: (2026)

IQ Test for LLMs: An Evaluation Framework for Uncovering Core Skills in LLMs
by: Maimon, Aviya, et al.
Published: (2025)

How Well Do LLMs Represent Values Across Cultures? Empirical Analysis of LLM Responses Based on Hofstede Cultural Dimensions
by: Kharchenko, Julia, et al.
Published: (2024)

Modality Collapse as Mismatched Decoding: Information-Theoretic Limits of Multimodal LLMs
by: Billa, Jayadev
Published: (2026)

A Thousand Words or An Image: Studying the Influence of Persona Modality in Multimodal LLMs
by: Broomfield, Julius, et al.
Published: (2025)

How Well Do Agentic Skills Work in the Wild: Benchmarking LLM Skill Usage in Realistic Settings
by: Liu, Yujian, et al.
Published: (2026)

Reading, Not Thinking: Understanding and Bridging the Modality Gap When Text Becomes Pixels in Multimodal LLMs
by: Sun, Kaiser, et al.
Published: (2026)

FutureOmni: Evaluating Future Forecasting from Omni-Modal Context for Multimodal LLMs
by: Chen, Qian, et al.
Published: (2026)

Traveling Across Languages: Benchmarking Cross-Lingual Consistency in Multimodal LLMs
by: Wang, Hao, et al.
Published: (2025)