The DVD framework converts a pretrained video diffusion model into a deterministic, single‑pass video depth estimator, eliminating random sampling artifacts, preserving geometric and semantic priors, and reaching zero‑shot state‑of‑the‑art performance with 163× less training data.
NOSE introduces a neural olfactory‑semantic embedding that unifies molecular structure, receptor sequences, and natural‑language odor descriptions into a continuous space, achieving state‑of‑the‑art results on eleven tasks and strong zero‑shot generalization for odor and receptor retrieval.
The RAAP framework decouples static contact point and dynamic action direction, using retrieval‑augmented inference to achieve zero‑sample, cross‑category fine‑grained robot manipulation with only a few training examples per task.
The article analyzes the Base‑to‑New generalization problem of CLIP‑based visual‑language models, explains why standard prompt tuning (CoOp) forgets base knowledge, and presents the KgCoOp framework that adds a knowledge‑guided loss to keep learned prompts close to hand‑crafted ones, dramatically improving unseen‑class performance while preserving efficiency.
The article analyzes why standard prompt tuning (CoOp) causes catastrophic forgetting in visual‑language models, introduces the KgCoOp framework that adds a knowledge‑guided loss to regularize prompts, and shows through extensive experiments on 11 benchmarks that KgCoOp improves unseen‑class accuracy, harmonic mean, and efficiency while discussing trade‑offs and limitations.
The article explains how contrastive learning replaces fixed‑category visual training with a relationship‑based approach, detailing the dual‑encoder architecture, cosine similarity loss, batch scaling, temperature control, zero‑shot capabilities, scalability from web data, and the method's strengths and limitations in modern multimodal AI.
TacRefineNet leverages high‑resolution tactile sensors, multimodal fusion of fingertip touch and proprioception, and a goal‑conditioned refinement network to achieve millimeter‑level grasp adjustments without vision or 3D models, demonstrating zero‑shot deployment and robust generalization across diverse automotive‑factory parts in both simulation and real‑world tests.
This article explains how Chain‑of‑Thought prompting and its variants—zero‑shot CoT, Thread of Thought, Tabular CoT, Analogical Prompting, and Step‑back Prompting—enable large language models to perform multi‑step reasoning by breaking problems into intermediate steps, with practical prompts, examples, and implementation details.
This study introduces EvoVLMA, an evolutionary vision-language model adaptation framework that automatically searches training-free VLM adaptation algorithms using a two-stage LLM-guided evolution, demonstrating superior performance—such as a 1.91 % accuracy gain on 8-shot image classification—and releasing the code publicly.
Jason Wei, a former Google Brain and OpenAI researcher now at Meta, has driven key advances in large language models—including chain‑of‑thought prompting, instruction tuning, emergent abilities, zero‑shot learning, and data augmentation—shaping both AI research paradigms and real‑world applications.
Nvidia introduces SeNaTra, a native‑segmentation vision transformer that replaces uniform down‑sampling with a content‑aware spatial grouping layer, delivering superior zero‑shot and supervised segmentation performance while cutting parameters and FLOPs compared with Swin Transformer and other backbones.
GraphGPT integrates large language models with graph neural networks by introducing graph tokens and instruction tuning, enabling zero‑shot graph learning for tasks such as node classification and link prediction, and demonstrates superior performance and generalization across supervised and zero‑shot benchmarks.
OpenGraph introduces a zero‑shot graph learning framework that unifies graph tokenization, a scalable transformer with efficient sampling, and LLM‑driven data augmentation, achieving superior cross‑dataset generalization on node classification and link prediction tasks, as demonstrated by extensive experiments.
The article reviews the Open Vocabulary Detection Contest organized by the Chinese Society of Image and Graphics and 360 AI Institute, describing the competition setup, dataset characteristics, and detailed winning approaches that combine Detic, CLIP, prompt learning, and multi‑stage pipelines to achieve strong few‑shot and zero‑shot object detection performance.
This article presents an in‑depth overview of Tencent's TexSmart natural‑language understanding system, highlighting its fine‑grained NER capabilities, knowledge‑base combination methods, remote‑supervision via similar entities, multi‑source zero‑shot fusion, experimental results, and practical insights from a recent NLP summit.
The article introduces TextToKnowledge, an open‑source Baidu platform that provides a unified Chinese term taxonomy (TermTree) and two annotation tools (WordTag and NPTag) to enable zero‑sample text labeling, term‑linking, and downstream knowledge‑mining applications for various NLP tasks.
This paper introduces MAIL, a model‑agnostic dual‑tower framework that uses a zero‑shot learning tower to generate virtual user behaviors for new users and an embedding‑based ranking tower, achieving 13‑15% CTR lift in large‑scale live‑stream recommendation at NetEase Cloud Music.
