GenEval

The 'GenEval' dataset/benchmark is used to evaluate the capabilities of multimodal large language models (MLLMs) in generating and refining images based on textual prompts.

Papers using GenEval (8)

Mural: Transferring LLM knowledge to image generation via Mixture-of-Transformers2026Cheers: Decoupling Patch Details from Semantic Representations Enables Unified Multimodal Comprehension and Generation2026i1: A Simple and Fully Open Recipe for Strong Text-to-Image Models2026Show, Don't Tell: Morphing Latent Reasoning into Image Generation2026AlphaGRPO: Unlocking Self-Reflective Multimodal Generation in UMMs via Decompositional Verifiable Reward2026Advantage Weighted Matching: Aligning RL with Pretraining in Diffusion Models2025CoRe^2: Collect, Reflect and Refine to Generate Better and Faster2025DraCo: Draft as CoT for Text-to-Image Preview and Rare Concept Generation2025