MM-JudgeBias benchmark shows that many MLLM judges neglect modalities and produce unstable evaluations under small input changes, based on tests of 26 models with over 1,800 samples.
The artbench dataset: Benchmarking generative models with artworks
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
verdicts
UNVERDICTED 6representative citing papers
MUCS uses mirrored unlearning and noise-consistent skew to outperform prior TDA methods for diffusion models on three datasets.
A technique for parametric stylistic control in latent diffusion models learns disentangled directions from synthetic datasets and applies them via guidance composition while preserving semantics.
HyCal mitigates Domain Gravity in cross-discipline imbalanced few-shot class-incremental learning by calibrating prototypes with complementary directional and covariance-aware distances on frozen CLIP embeddings.
Insert In Style is a zero-shot framework that disentangles identity, style, and composition via multi-stage training, masked attention, and prior preservation to enable harmonious cross-domain object insertion in images.
Hybrid knowledge graph embeddings fused with vision transformer features outperform standard techniques on abstract concept classification by integrating situated perceptual knowledge from a new cultural image resource.
citing papers explorer
-
MM-JudgeBias: A Benchmark for Evaluating Compositional Biases in MLLM-as-a-Judge
MM-JudgeBias benchmark shows that many MLLM judges neglect modalities and produce unstable evaluations under small input changes, based on tests of 26 models with over 1,800 samples.
-
Training data attribution in diffusion models via mirrored unlearning and noise-consistent skew
MUCS uses mirrored unlearning and noise-consistent skew to outperform prior TDA methods for diffusion models on three datasets.
-
Stylistic Attribute Control in Latent Diffusion Models
A technique for parametric stylistic control in latent diffusion models learns disentangled directions from synthetic datasets and applies them via guidance composition while preserving semantics.
-
HyCal: A Training-Free Prototype Calibration Method for Cross-Discipline Few-Shot Class-Incremental Learning
HyCal mitigates Domain Gravity in cross-discipline imbalanced few-shot class-incremental learning by calibrating prototypes with complementary directional and covariance-aware distances on frozen CLIP embeddings.
-
Insert In Style: A Zero-Shot Generative Framework for Harmonious Cross-Domain Object Composition
Insert In Style is a zero-shot framework that disentangles identity, style, and composition via multi-stage training, masked attention, and prior preservation to enable harmonious cross-domain object insertion in images.
-
Stitching Gaps: Fusing Situated Perceptual Knowledge with Vision Transformers for High-Level Image Classification
Hybrid knowledge graph embeddings fused with vision transformer features outperform standard techniques on abstract concept classification by integrating situated perceptual knowledge from a new cultural image resource.