ReMoA: Recursive Modular Adaptation via Sparse Latent Refinement in Multimodal Diffusion Models
cs.CV, cs.AI
Submitted: 2026-04-28
Updated: 2026-11-04
Terminology
Sources
- Rationale-Augmented Ensembles in Language Models
- Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters
- Fast and Slow Learning of Recurrent Independent Mechanisms
- LLaDA-MoE: A Sparse MoE Diffusion Language Model
- Dynamic Diffusion Transformer
- Compete and Compose: Learning Independent Mechanisms for Modular World Models
- Denoising Diffusion Implicit Models
- PixArt-$\alpha$: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis
- Coordination Among Neural Modules Through a Shared Global Workspace
- Consciousness in Artificial Intelligence: Insights from the Science of Consciousness
- DyDiT++: Diffusion Transformers with Timestep and Spatial Dynamics for Efficient Visual Generation
- PonderNet: Learning to Ponder
- Less is More: Recursive Reasoning with Tiny Networks
- Hierarchical Reasoning Model
- Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach
- LaRe: Latent Refocusing for Multimodal Reasoning
- Multimodal Chain of Continuous Thought for Latent-Space Reasoning in Vision-Language Models
- LoRA: Low-Rank Adaptation of Large Language Models
- ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment
- VSP: Assessing the dual challenges of perception and reasoning in spatial planning tasks for VLMs
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs