VITRON is a pixel-level visual large-scale language model (LLM) jointly developed by Skywork AI, the National University of Singapore, and Nanyang Technological University. It can comprehensively understand and process static images and dynamic videos, enabling image and video comprehension...
n8n is an open-source automated workflow management system that provides a low-code platform that allows users to create complex workflows via drag and drop, without writing any code. n8n supports integration with over 400 applications and services, including AI...
SoulChat 2.0 is a digital twin language model for psychological counselors, developed by the Future Technology Institute of South China University of Technology - Guangdong Provincial Key Laboratory of Digital Twins, based on the SoulChat 1.0 model. It defines for the first time the digital twin language of a specific psychological counselor...
SAC-KG is a general framework for automatically constructing domain knowledge graphs based on large-scale language models (LLMs). It consists of three main components: a generator, a validator, and a pruner, and can automatically generate domain-specific knowledge graphs from a raw domain corpus...
TryOffAnyone is an AI tool that transforms photos of people wearing clothing into flat-lay clothing display images. Based on advanced Latent Diffusion Models technology, it automatically identifies and extracts clothing areas from photos, and then...
LongDocURL is a multimodal long document understanding benchmark dataset jointly released by the Institute of Automation, Chinese Academy of Sciences, and Alibaba's Taobao Tmall Group. It focuses on evaluating models' performance in handling long documents, complex elements, and diverse tasks...
DeepSeek Artifacts is a free AI programming tool from Hugging Face that generates front-end code for React and Tailwind CSS. Based on DeepSeek V3, its primary goal is to build public datasets of front-end code...
Jina Reader is an open-source tool from Jina AI that converts HTML web page content into plain text format suitable for processing by large language models (LLMs). Users simply add a specific prefix to the URL, i.e....
Memory Layers, introduced by Meta, are a method for adding extra parameters to a model using a trainable key-value lookup mechanism without increasing the number of floating-point operations (FLOPs). They supplement computationally intensive feedforward layers with sparse activation, providing...
Cognita is an open-source, modular RAG (Retrieval Augmented Generation) framework that helps developers build modular, easily scalable, and deployable production-grade applications. Cognita uses Langchain/LlamaInd...
LAM is a new artificial intelligence model launched by Microsoft called "Large Action Model" (LAM for short). Unlike traditional language models, LAM can autonomously operate Windows programs and perform real-world tasks. It can understand text...
smolagents is a lightweight open-source library from Hugging Face that simplifies the process of building intelligent agents. Its minimal code size (approximately 1000 lines) and intuitive API design enable developers to quickly build and deploy intelligent agents. smo...
REEF (Representation Encoding Fingerprints) is a fingerprinting technique used for large language models (LLMs). It generates a unique fingerprint for each model by embedding specific encoded information during model training...
Inf-DiT, developed by Tsinghua University and Zhipu AI, is an image upsampling method based on a diffusion model that can generate ultra-high resolution images. Inf-DiT introduces a unidirectional block attention mechanism (UniBA), reducing the space complexity of the generation process from...
ImBD (Imitate Before Detect) is a method developed by Fudan University, South China University of Technology, Wuhan University, Fenzi AI, and others for detecting machine-edited text. It first mimics the text generation process of large language models (LLMs)...
SocraticLM is an innovative AI model for teaching mathematics using the Socratic method. Developed jointly by the University of Science and Technology of China (USTC) and iFlytek, it employs a "Thought-Provoking" teaching paradigm, guiding learning through questioning...
VMix is an innovative plug-and-play aesthetic adapter that enhances the aesthetic quality of images generated by text-to-image diffusion models. By decoupling the content description and aesthetic description in the input text prompts, it integrates fine-grained aesthetic tags (such as color, ...
smoltalk-chinese is an open-source synthetic dataset from OpenCSG specifically designed for large-scale Chinese language models (LLMs). This dataset contains over 700,000 synthetic data entries, covering information retrieval, reasoning, planning, editing, and more...
RLCM (Reinforcement Learning for Consistency Model) is a framework introduced by Cornell University for optimizing text-to-image generation models. It uses reinforcement learning methods to fine-tune consistency models to adapt them to specific tasks...
LLM2LLM is an innovative iterative data augmentation strategy that improves the performance of large language models (LLMs) in data-scarce scenarios. The method enhances the training performance of student models by generating synthetic data based on a powerful teacher model...
CodeElo is a benchmark tool for evaluating the code generation capabilities of Large Language Models (LLMs) at the programming competition level. It measures the programming skill level of LLMs by comparing them to the Elo rating system for human programmers. ...
FlexRAG is an innovative retrieval-enhanced generation (RAG) framework that addresses the high computational cost and insufficient generation quality of traditional RAG systems when handling long contexts. It achieves this by compressing the retrieved context information into a compact...
GeneralDyG is a general dynamic graph anomaly detection method proposed by a research team at Nanyang Technological University. It addresses the anomaly detection problem of dynamic graph data in fields such as social networks, e-commerce, and cybersecurity. It utilizes time-based ego-graph sampling...
OpenHands is an AI programming tool that supports multi-agent collaboration to improve development efficiency and reduce developers' coding workload. OpenHands enables interaction with the environment through writing code, interacting with command lines, and browsing web pages.