LoRA adapters fix collapsed visual CLS token attention in CLIP for superior cross-domain few-shot learning, and the new Semantic Probe framework revives prompt methods to reach state-of-the-art on four benchmarks.
Two effects, one trigger: on the modality gap, object bias, and information imbalance in contrastive vision-language representation learning
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
fields
cs.CV 6years
2026 6verdicts
UNVERDICTED 6roles
background 2polarities
background 2representative citing papers
GPUA learns an orthogonal mapping from VFM to VLM feature space to preserve geometry and improve cross-model compatibility for zero-shot recognition and segmentation.
LoMo is a lightweight data curation technique that locally substitutes text with images in prompts to enforce cross-modal invariance, yielding 2.67-2.82 point gains over standard SFT on two VLMs across 13 benchmarks.
VLMs fail at counting because visual evidence degrades in later language layers, and a lightweight Modality Attention Share intervention can encourage better use of image information during answer generation.
VL-SAM-v3 retrieves visual prototypes from memory to generate sparse spatial and dense contextual priors that refine detection prompts, yielding gains on rare categories in LVIS for both open-vocabulary and open-ended settings.
A pre-trained alignment model is adapted for multimodal coreference resolution via similarity aggregation and evidence theory, reporting gains on the CIN benchmark over prior dedicated methods and VLLMs.
citing papers explorer
-
Reviving In-domain Fine-tuning Methods for Source-Free Cross-domain Few-shot Learning
LoRA adapters fix collapsed visual CLS token attention in CLIP for superior cross-domain few-shot learning, and the new Semantic Probe framework revives prompt methods to reach state-of-the-art on four benchmarks.
-
Geometry-Preserving Unsupervised Alignment for Heterogeneous Foundation Models
GPUA learns an orthogonal mapping from VFM to VLM feature space to preserve geometry and improve cross-model compatibility for zero-shot recognition and segmentation.
-
LoMo: Local Modality Substitution for Deeper Vision-Language Fusion
LoMo is a lightweight data curation technique that locally substitutes text with images in prompts to enforce cross-modal invariance, yielding 2.67-2.82 point gains over standard SFT on two VLMs across 13 benchmarks.
-
Counting to Four is still a Chore for VLMs
VLMs fail at counting because visual evidence degrades in later language layers, and a lightweight Modality Attention Share intervention can encourage better use of image information during answer generation.
-
VL-SAM-v3: Memory-Guided Visual Priors for Open-World Object Detection
VL-SAM-v3 retrieves visual prototypes from memory to generate sparse spatial and dense contextual priors that refine detection prompts, yielding gains on rare categories in LVIS for both open-vocabulary and open-ended settings.
-
Plug-and-Adapt: Multimodal Coreference Resolution at First Sight with a Pretrained Alignment Model
A pre-trained alignment model is adapted for multimodal coreference resolution via similarity aggregation and evidence theory, reporting gains on the CIN benchmark over prior dedicated methods and VLLMs.