Cover Image

From Drop-off to Recovery: A Mechanistic Analysis of Segmentation in MLLMs

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Wu, Boyong, Kim, Sanghwan, Akata, Zeynep
Format:	Preprint
Published:	2026
Subjects:	Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

FINER: MLLMs Hallucinate under Fine-grained Negative Queries
by: Xiao, Rui, et al.
Published: (2026)

COSMOS: Cross-Modality Self-Distillation for Vision Language Pre-training
by: Kim, Sanghwan, et al.
Published: (2024)

Improving Intervention Efficacy via Concept Realignment in Concept Bottleneck Models
by: Singhi, Nishad, et al.
Published: (2024)

FLAIR: VLM with Fine-grained Language-informed Image Representations
by: Xiao, Rui, et al.
Published: (2024)

SUB: Benchmarking CBM Generalization via Synthetic Attribute Substitutions
by: Bader, Jessica, et al.
Published: (2025)

Training-free Uncertainty Guidance for Complex Visual Tasks with MLLMs
by: Kim, Sanghwan, et al.
Published: (2025)

Concept-Guided Interpretability via Neural Chunking
by: Wu, Shuchen, et al.
Published: (2025)

The Latent Color Subspace: Emergent Order in High-Dimensional Chaos
by: Pach, Mateusz, et al.
Published: (2026)

Time Series Representations for Classification Lie Hidden in Pretrained Vision Transformers
by: Roschmann, Simon, et al.
Published: (2025)

Sparse Autoencoders Learn Monosemantic Features in Vision-Language Models
by: Pach, Mateusz, et al.
Published: (2025)

Stitch: Training-Free Position Control in Multimodal Diffusion Transformers
by: Bader, Jessica, et al.
Published: (2025)

Unbalancedness in Neural Monge Maps Improves Unpaired Domain Translation
by: Eyring, Luca, et al.
Published: (2023)

A More Word-like Image Tokenization for MLLMs
by: Lee, Hyun, et al.
Published: (2026)

Sparse Autoencoders are Topic Models
by: Girrbach, Leander, et al.
Published: (2025)

AttentionDrop: A Novel Regularization Method for Transformer Models
by: Baig, Mirza Samad Ahmed, et al.
Published: (2025)

RAR: Retrieving And Ranking Augmented MLLMs for Visual Recognition
by: Liu, Ziyu, et al.
Published: (2024)

HueManity: Probing Fine-Grained Visual Perception in MLLMs
by: Grover, Rynaa, et al.
Published: (2025)

PerNodeDrop: A Method Balancing Specialized Subnets and Regularization in Deep Neural Networks
by: Omathil, Gelesh G, et al.
Published: (2025)

APO: Enhancing Reasoning Ability of MLLMs via Asymmetric Policy Optimization
by: Hong, Minjie, et al.
Published: (2025)

Simple Drop-in LoRA Conditioning on Attention Layers Will Improve Your Diffusion Model
by: Choi, Joo Young, et al.
Published: (2024)

STARFISH: faST Accuracy Recovery in pruned networks From Internal State Healing
by: Maon, Shir, et al.
Published: (2026)

Observe-R1: Unlocking Reasoning Abilities of MLLMs with Dynamic Progressive Reinforcement Learning
by: Guo, Zirun, et al.
Published: (2025)

MileBench: Benchmarking MLLMs in Long Context
by: Song, Dingjie, et al.
Published: (2024)

MLLMs-Augmented Visual-Language Representation Learning
by: Liu, Yanqing, et al.
Published: (2023)

InfoTok: Information-Theoretic Regularization for Capacity-Constrained Shared Visual Tokenization in Unified MLLMs
by: Tang, Lv, et al.
Published: (2026)

Perception-R1: Advancing Multimodal Reasoning Capabilities of MLLMs via Visual Perception Reward
by: Xiao, Tong, et al.
Published: (2025)

Causal Unsupervised Semantic Segmentation
by: Kim, Junho, et al.
Published: (2023)

DeLoRA: Decoupling Angles and Strength in Low-rank Adaptation
by: Bini, Massimo, et al.
Published: (2025)

SynRES: Towards Referring Expression Segmentation in the Wild via Synthetic Data
by: Kim, Dong-Hee, et al.
Published: (2025)

Towards a Mechanistic Explanation of Diffusion Model Generalization
by: Niedoba, Matthew, et al.
Published: (2024)

Prisma: An Open Source Toolkit for Mechanistic Interpretability in Vision and Video
by: Joseph, Sonia, et al.
Published: (2025)

Concept-Based Mechanistic Interpretability Using Structured Knowledge Graphs
by: Chorna, Sofiia, et al.
Published: (2025)

Segment Any Vehicle: Semantic and Visual Context Driven SAM and A Benchmark
by: Wang, Xiao, et al.
Published: (2025)

FOCUS: Towards Universal Foreground Segmentation
by: You, Zuyao, et al.
Published: (2025)

OTSeg: Multi-prompt Sinkhorn Attention for Zero-Shot Semantic Segmentation
by: Kim, Kwanyoung, et al.
Published: (2024)

R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce Learning
by: Yang, Qi, et al.
Published: (2025)

SALVE: Sparse Autoencoder-Latent Vector Editing for Mechanistic Control of Neural Networks
by: Flovik, Vegard
Published: (2025)

Tables as Texts or Images: Evaluating the Table Reasoning Ability of LLMs and MLLMs
by: Deng, Naihao, et al.
Published: (2024)

Model Merging in LLMs, MLLMs, and Beyond: Methods, Theories, Applications and Opportunities
by: Yang, Enneng, et al.
Published: (2024)

M2-Reasoning: Empowering MLLMs with Unified General and Spatial Reasoning
by: AI, Inclusion, et al.
Published: (2025)