Imagen de Portada

Training-Free Mitigation of Language Reasoning Degradation After Multimodal Instruction Tuning

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Ratzlaff, Neale, Luo, Man, Su, Xin, Lal, Vasudev, Howard, Phillip
Formato:	Preprint
Publicado:	2024
Materias:	Computer Vision and Pattern Recognition Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Pruning the Paradox: How CLIP's Most Informative Heads Enhance Performance While Amplifying Bias
por: Madasu, Avinash, et al.
Publicado: (2025)

DPO Learning with LLMs-Judge Signal for Computer Use Agents
por: Luo, Man, et al.
Publicado: (2025)

Cultural Awareness in Vision-Language Models: A Cross-Country Exploration
por: Madasu, Avinash, et al.
Publicado: (2025)

SocialCounterfactuals: Probing and Mitigating Intersectional Social Biases in Vision-Language Models with Counterfactual Examples
por: Howard, Phillip, et al.
Publicado: (2023)

Quantifying and Enabling the Interpretability of CLIP-like Models
por: Madasu, Avinash, et al.
Publicado: (2024)

Debiasing Large Vision-Language Models by Ablating Protected Attribute Representations
por: Ratzlaff, Neale, et al.
Publicado: (2024)

SK-VQA: Synthetic Knowledge Generation at Scale for Training Context-Augmented Multimodal LLMs
por: Su, Xin, et al.
Publicado: (2024)

Probing the Representational Power of Sparse Autoencoders in Vision Models
por: Olson, Matthew Lyle, et al.
Publicado: (2025)

Debias your Large Multi-Modal Model at Test-Time via Non-Contrastive Visual Attribute Steering
por: Ratzlaff, Neale, et al.
Publicado: (2024)

Analyzing Hierarchical Structure in Vision Models with Sparse Autoencoders
por: Olson, Matthew Lyle, et al.
Publicado: (2025)

FiVL: A Framework for Improved Vision-Language Alignment through the Lens of Training, Evaluation and Explainability
por: Aflalo, Estelle, et al.
Publicado: (2024)

ICSVR: Investigating Compositional and Syntactic Understanding in Video Retrieval Models
por: Madasu, Avinash, et al.
Publicado: (2023)

ScalSelect: Scalable Training-Free Multimodal Data Selection for Efficient Visual Instruction Tuning
por: Wu, Changti, et al.
Publicado: (2026)

Cross-Cultural Value Awareness in Large Vision-Language Models
por: Howard, Phillip, et al.
Publicado: (2026)

LVLM-Compress-Bench: Benchmarking the Broader Impact of Large Vision-Language Model Compression
por: Kundu, Souvik, et al.
Publicado: (2025)

Seg-Agent: Test-Time Multimodal Reasoning for Training-Free Language-Guided Segmentation
por: Hao, Chao, et al.
Publicado: (2026)

mEOL: Training-Free Instruction-Guided Multimodal Embedder for Vector Graphics and Image Retrieval
por: Kim, Kyeong Seon, et al.
Publicado: (2026)

Prescribing the Right Remedy: Mitigating Hallucinations in Large Vision-Language Models via Targeted Instruction Tuning
por: Hu, Rui, et al.
Publicado: (2024)

MCITlib: Multimodal Continual Instruction Tuning Library and Benchmark
por: Guo, Haiyang, et al.
Publicado: (2025)

PathInsight: Instruction Tuning of Multimodal Datasets and Models for Intelligence Assisted Diagnosis in Histopathology
por: Wu, Xiaomin, et al.
Publicado: (2024)

Mitigating the Reasoning Tax in Vision-Language Fine-Tuning with Input-Adaptive Depth Aggregation
por: Ren, Yiming, et al.
Publicado: (2026)

MedHallTune: An Instruction-Tuning Benchmark for Mitigating Medical Hallucination in Vision-Language Models
por: Yan, Qiao, et al.
Publicado: (2025)

Advancing Multimodal Large Language Models in Chart Question Answering with Visualization-Referenced Instruction Tuning
por: Zeng, Xingchen, et al.
Publicado: (2024)

Steering Large Language Models to Evaluate and Amplify Creativity
por: Olson, Matthew Lyle, et al.
Publicado: (2024)

Instruction-tuned Self-Questioning Framework for Multimodal Reasoning
por: Jang, You-Won, et al.
Publicado: (2025)

Text-Guided Layer Fusion Mitigates Hallucination in Multimodal LLMs
por: Lin, Chenchen, et al.
Publicado: (2026)

Refining Text-to-Image Generation: Towards Accurate Training-Free Glyph-Enhanced Image Generation
por: Lakhanpal, Sanyam, et al.
Publicado: (2024)

ClueTracer: Question-to-Vision Clue Tracing for Training-Free Hallucination Suppression in Multimodal Reasoning
por: Xi, Gongli, et al.
Publicado: (2026)

BAMI: Training-Free Bias Mitigation in GUI Grounding
por: Zhang, Borui, et al.
Publicado: (2026)

VisNec: Measuring and Leveraging Visual Necessity for Multimodal Instruction Tuning
por: Dong, Mingkang, et al.
Publicado: (2026)

See&Trek: Training-Free Spatial Prompting for Multimodal Large Language Model
por: Li, Pengteng, et al.
Publicado: (2025)

Is 'Right' Right? Enhancing Object Orientation Understanding in Multimodal Large Language Models through Egocentric Instruction Tuning
por: Jung, Ji Hyeok, et al.
Publicado: (2024)

Reinforcement Fine-Tuning Powers Reasoning Capability of Multimodal Large Language Models
por: Sun, Haoyuan, et al.
Publicado: (2025)

Chain-of-Thought Degrades Visual Spatial Reasoning Capabilities of Multimodal LLMs
por: Kancheti, Sai Srinivas, et al.
Publicado: (2026)

VISTA: Mitigating Semantic Inertia in Video-LLMs via Training-Free Dynamic Chain-of-Thought Routing
por: Jin, Hongbo, et al.
Publicado: (2025)

Bridging the Dynamic Perception Gap: Training-Free Draft Chain-of-Thought for Dynamic Multimodal Spatial Reasoning
por: Ou, Siqu, et al.
Publicado: (2025)

Uncovering Bias in Large Vision-Language Models with Counterfactuals
por: Howard, Phillip, et al.
Publicado: (2024)

EMO-R3: Reflective Reinforcement Learning for Emotional Reasoning in Multimodal Large Language Models
por: Fang, Yiyang, et al.
Publicado: (2026)

Deeper Thought, Weaker Aim: Understanding and Mitigating Perceptual Impairment during Reasoning in Multimodal Large Language Models
por: Peng, Ruiying, et al.
Publicado: (2026)

Aligning Vision to Language: Annotation-Free Multimodal Knowledge Graph Construction for Enhanced LLMs Reasoning
por: Liu, Junming, et al.
Publicado: (2025)