XTC-Bench reveals that strong performance on generation or understanding tasks in unified multimodal models does not guarantee cross-task semantic consistency, which instead depends on how tightly coupled the learning objectives are across modalities.
Vision as a dialect: Unifying visual understanding and generation via text-aligned representations.arXiv preprint arXiv:2506.18898
12 Pith papers cite this work. Polarity classification is still indexing.
12
Pith papers citing it
citation-role summary
baseline 2
background 1
citation-polarity summary
years
2026 12representative citing papers
Hierarchical Binary Quantization plus global refinement AR yields 0.56 rFID reconstruction and 1.81 gFID class-conditional generation on ImageNet, with competitive T2I/T2V at 2B scale.
HYDRA-X presents the first unified multimodal model using a single ViT for holistic image-video tokenization, with ablations on attention and compression plus a latent-level editing improvement.
Residual Decoder Adapter upgrades existing visual tokenizers for AR image models by adding a paired codebook and pixel-space residual branch, raising OCR accuracy on text rendering benchmarks from ~12-24% to ~36-58%.
MUSE decouples reconstruction and semantic learning in visual tokenization via topological orthogonality, yielding SOTA generation quality and improved semantic performance over its teacher model.
Tuna-2 shows that direct pixel embeddings can replace vision encoders in unified multimodal models, achieving competitive generation and stronger understanding at scale.
By requiring and using highly discriminative LLM text features, the work enables the first effective one-step text-conditioned image generation with MeanFlow.
InfoTok uses mutual information constraints to regularize shared visual tokenization in unified MLLMs, improving both understanding and generation performance without extra training data.
Vega unifies video understanding and generation via shared vocabulary and hybrid autoregressive-diffusion architecture, reporting strong results on VBench and VideoMME.
SPAR introduces a semantic-pixel self-alignment tokenizer and dynamic token routing to create a unified multimodal model that performs both understanding and generation at claimed state-of-the-art levels.
ARM is a 7B autoregressive multimodal model with a unified discrete visual tokenizer and RL that performs image understanding, generation, and editing while showing cross-task synergy from preference optimization.
Visual generation models are evolving from passive renderers to interactive agentic world modelers, but current systems lack spatial reasoning, temporal consistency, and causal understanding, with evaluations overemphasizing perceptual quality.
citing papers explorer
-
Beyond Accuracy: Benchmarking Cross-Task Consistency in Unified Multimodal Models
XTC-Bench reveals that strong performance on generation or understanding tasks in unified multimodal models does not guarantee cross-task semantic consistency, which instead depends on how tightly coupled the learning objectives are across modalities.
-
Generative Refinement Networks for Visual Synthesis
Hierarchical Binary Quantization plus global refinement AR yields 0.56 rFID reconstruction and 1.81 gFID class-conditional generation on ImageNet, with competitive T2I/T2V at 2B scale.
-
HYDRA-X: Native Unified Multimodal Models with Holistic Visual Tokenizers
HYDRA-X presents the first unified multimodal model using a single ViT for holistic image-video tokenization, with ablations on attention and compression plus a latent-level editing improvement.
-
Residual Decoder Adapter: ID-Preserving Tokenizer Adaption for Autoregressive Text Rendering
Residual Decoder Adapter upgrades existing visual tokenizers for AR image models by adding a paired codebook and pixel-space residual branch, raising OCR accuracy on text rendering benchmarks from ~12-24% to ~36-58%.
-
MUSE: Resolving Manifold Misalignment in Visual Tokenization via Topological Orthogonality
MUSE decouples reconstruction and semantic learning in visual tokenization via topological orthogonality, yielding SOTA generation quality and improved semantic performance over its teacher model.
-
Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generation
Tuna-2 shows that direct pixel embeddings can replace vision encoders in unified multimodal models, achieving competitive generation and stronger understanding at scale.
-
Extending One-Step Image Generation from Class Labels to Text via Discriminative Text Representation
By requiring and using highly discriminative LLM text features, the work enables the first effective one-step text-conditioned image generation with MeanFlow.
-
InfoTok: Information-Theoretic Regularization for Capacity-Constrained Shared Visual Tokenization in Unified MLLMs
InfoTok uses mutual information constraints to regularize shared visual tokenization in unified MLLMs, improving both understanding and generation performance without extra training data.
-
Bridging Video Understanding and Generation in a Unified Framework
Vega unifies video understanding and generation via shared vocabulary and hybrid autoregressive-diffusion architecture, reporting strong results on VBench and VideoMME.
-
SPAR: Semantic-Pixel Self-Alignment and Adaptive Routing for Unified Multimodal Models
SPAR introduces a semantic-pixel self-alignment tokenizer and dynamic token routing to create a unified multimodal model that performs both understanding and generation at claimed state-of-the-art levels.
-
ARM: An AutoRegressive Large Multimodal Model with Unified Discrete Representations
ARM is a 7B autoregressive multimodal model with a unified discrete visual tokenizer and RL that performs image understanding, generation, and editing while showing cross-task synergy from preference optimization.
-
Visual Generation in the New Era: An Evolution from Atomic Mapping to Agentic World Modeling
Visual generation models are evolving from passive renderers to interactive agentic world modelers, but current systems lack spatial reasoning, temporal consistency, and causal understanding, with evaluations overemphasizing perceptual quality.