arXiv AI
Jul 28
What CLIP Knows but Cannot Say: Recovering Negation from Frozen Intermediate Features
arXiv:2607. 23271v1 Announce Type: cross Abstract: Contrastive vision-language models such as CLIP map semantically opposite phrases (e.
Know-Thy-Not: Training-Free Targeted Negation in Dual-Encoder Vision-Language Models
Related stories
arXiv AI
Sep 15
Not all Negation Cues are Equal: Affixal Negations Yield Better Negation Understanding
The paper introduces NegCue, a large-scale dataset of 1.8 million samples that includes single-word, multi-word, and affixal negation cues, totaling over 200 unique forms. The authors pre-train encoder-only language models and large language models on this dataset to study how different negation types influence understanding. Experiments on five downstream benchmarks reveal that affixal negations provide the most significant performance gains, whereas single-word negations yield modest improvements, and that additional pre-training benefits both model types.
arXiv Machine Learning
Jun 30
Contrastive vision-language learning with paraphrasing and negation
arXiv:2511. 16527v2 Announce Type: replace-cross Abstract: Contrastive vision-language models continue to be the dominant approach for image-text retrieval.
arXiv Computer Vision
Sep 15
HANCLIP: A Family of Hyperbolic Angular Negation Vision Language Models
arXiv:2606.23843v2 Announce Type: replace Abstract: Vision-language models (VLMs) achieve strong cross-modal alignment but remain brittle to negation, often relying on shallow word associations rathe...
arXiv Computer Vision
Oct 9
Image Recognition with Vision and Language Embeddings of VLMs
The paper evaluates how vision-language models (VLMs) perform in image classification tasks, comparing language-guided and vision-only approaches across various dual-encoder VLMs such as SigLIP 2 and RADIOv2.5 on ImageNet-1k and its label-corrected variant. It identifies key factors influencing accuracy—including prompt design, class diversity, k‑NN neighbors, and reference set size—and demonstrates that language and vision provide complementary strengths. A simple, learning‑free fusion method based on per‑class precision is proposed to further improve classification performance.
arXiv Machine Learning
Aug 18
FuseLIP: Multimodal Embeddings via Early Fusion of Discrete Tokens
arXiv:2506. 03096v2 Announce Type: replace-cross Abstract: Contrastive language-image pre-training aligns features of text-image pairs in a common latent space via distinct encoders for each modality.
arXiv AI
Aug 5
From Generator to Embedder: Harnessing Innate Abilities of Multimodal LLMs via Building Zero-Shot Discriminative Embedding Model
arXiv:2508. 00955v3 Announce Type: replace-cross Abstract: Adapting generative Multimodal Large Language Models (MLLMs) into universal embedding models typically demands resource-intensive contrastive pre-training, while traditional hard negative mining methods suffer from severe false negative contamination.
arXiv Machine Learning
Jul 28
Similarity Is Not Logic: Factored Inference for Dual-Encoder Vision-Language Models
arXiv:2607. 23052v1 Announce Type: cross Abstract: Dual-encoder vision-language models (VLMs) expose a similarity interface that enables zero-shot retrieval but fails compositional constraints: queries like "umbrella and no person" retrieve images containing both, even when concept detection is reliable.
arXiv AI
Oct 9
Discovering Global False Negatives On the Fly for Self-supervised Contrastive Learning
The paper introduces GloFND, an optimization-based method that learns a threshold for each anchor image during training to automatically detect false negative pairs in self-supervised contrastive learning. Unlike prior approaches that only consider false negatives within a mini-batch, GloFND globally identifies them across the entire dataset while keeping per-iteration computation independent of dataset size. Experiments on image and image-text data show the method’s effectiveness, and the implementation is publicly available.
Hugging Face Trending Papers
Jul 2
FlowCIR: Semantic Transport via Flow Matching for Zero-Shot Composed Image Retrieval
Zero-shot composed image retrieval (ZS-CIR) aims to retrieve a target image by editing a reference image with a natural-language instruction, without relying on domain-specific annotated triplets. Most existing ZS-CIR methods rely on textual inversion to translate the reference image into pseudo-text tokens and then compose them with the instruction via simple concatenation in the text space, which can be lossy and brittle for fine-grained semantics.
arXiv AI
Jul 15
The Sound of Absence: Audio-Language Embedding Models Struggle with Negation
arXiv:2607. 12290v1 Announce Type: cross Abstract: Audio-language embedding models such as CLAP are widely evaluated on matching present sound events, but rarely on negation.
arXiv Computer Vision
Oct 2
Skeleton-and-Strategy Prompting: Training-Free Negation Understanding for Vision-Language Models
arXiv:2610.01180v1 Announce Type: new Abstract: Despite the strong performance of Vision-Language Models (VLMs) on a wide range of visual question answering (VQA) tasks, these models consistently str...