Tag
Meta-CLIP introduces a metadata-conditioned contrastive learning framework that reweights cross-modal negatives using DICOM acquisition fields, improving zero-shot disease classification on CT-RATE and several CXR benchmarks over CT-CLIP.
Jeff is a fine-tuned version of Qwen3.5 and Gemma 4 for zero-shot classification, capable of running quickly on local devices with performance approaching Jev on certain benchmarks.
The Jeff models are ultra-fast, open-weight decision models trained locally on hardware like RTX PRO 6000, achieving competitive benchmarks against larger models like Jev with sub-30ms inference speeds.
A tweet questions whether 'decision model' is a new term for zero-shot classification in AI.
GeoTTER presents a novel optimal transport framework for zero-shot classification that incorporates local geometric structure and clustering-guided cost components, achieving significant performance improvements over baseline methods.
This paper proposes a zero-shot detection pipeline to characterize human-likeness in AI-generated poetry, aiming to identify attributes that distinguish human from machine poems and improve detection methods.
This paper uses a pre-trained LLM with zero-shot classification to analyze approximately 20 million Twitch chat messages across seven game genres, finding that 2.4% of messages are toxic, with MOBA games having the highest rate (3.2%) and sports games the lowest (2%). The study also identifies significant differences in toxicity distributions across individual games within the same genre.