DeerFlow is a byte jump open source deep research framework designed for efficiently accomplishing complex research tasks. The framework integrates multiple language models (e.g. Qwen) and tools (e.g. search engines , crawlers , Python execution environment ) , can quickly generate high-quality research reports , podcasts ...

Dolphin is a byte jump open source lightweight document parsing large model , with 322M parameters , small size and fast running speed . The model is based on a two-stage parsing approach , based on page-level layout analysis to identify the elements in the document ( such as headings , tables, formulas , etc.) , and then the content of each element parsed ...

PromptPilot is an AI prompt optimization platform launched by Volcano Engine under ByteDance. The platform deeply analyzes the user's intention and transforms vague ideas into professional instructions that can be precisely executed by AI, ensuring that the model outputs high-quality results. Based on interactive guidance, PromptP...

Kimi-Dev is an open source code model from Moonshot AI designed for software engineering with 72B parameters. The model has a powerful BugFixer feature that can automatically locate and fix code errors, and the model provides test code generation...

HIRING.STUDIO is an AI recruitment interview assistance tool that helps recruitment teams efficiently create job-specific interview questions. Based on AI technology, the tool generates interview questions, suggests follow-up questions, and develops evaluation criteria for answers of varying quality. The tool supports the STAR method...

QVQ-Max is a state-of-the-art visual reasoning model introduced by Alitonix, which is an upgraded version of QVQ-72B-Preview. QVQ-Max is an advanced visual reasoning model that can "read" images and video content and combine the information for analysis, reasoning and problem solving.QVQ-Max's main functions include image parsing, video analysis...

MiniMax-M1 is an open source inference model from the MiniMax team, based on a combination of the Mixed Expert Architecture (MoE) and the Lightning Attention mechanism, with 456 billion total parameters. The model supports 100...

n8n is a powerful open source automation tool that supports the creation of workflows through a visual interface that connects various applications and services to automate tasks. n8n brings together a wide range of useful workflow templates that cover everything from simple data processing to complex business process automation.

Agent Zero is an open-source artificial intelligence framework to create general-purpose, highly customizable intelligent assistants. Through dynamic learning and evolution, it is able to handle a wide range of tasks, with persistent memory capabilities that remember previous experiences and solutions to accomplish subsequent tasks more efficiently.

Klic Studio (formerly Krillin AI) is an AI-based video translation, dubbing and voice cloning tool designed for video creators and content exporters. It supports one-click deployment of the whole process, which can complete the video from download to finished output in one click, and is adapted to Jittery, Xiaohongshu, B...

Resume Builder is a professional AI resume builder tool designed for job seekers. The tool combines AI technology to provide keywords and writing suggestions based on job descriptions to help users write resumes that precisely match hiring requirements. The platform has more than 25 customizable resume templates and super...

ConverzAI is a recruiting platform based on AI technology that optimizes the hiring process. The platform is based on deploying virtual recruiters (AI Agents) to automatically collect candidate information from the Application Tracking System (ATS) and perform initial screening. The platform uses multiple channels such as phone, email, SMS...

Mercor is an AI-driven platform for global recruiting that streamlines the job search and hiring process. Job seekers upload their resumes to complete a 20-minute AI interview, and Mercor matches the user's information to thousands of job requests from companies around the world to quickly find remote job opportunities.Mercor core features...

Clacky AI is a cloud-based development environment (CDE) for developers dedicated to creating L3-level AI programming assistants. The tool supports multiple programming languages, including Python, Node.js, Golang, Ruby, and Java, and can help developers quickly put...

MagicTryOn is an advanced video virtual try-on framework launched by the School of Computer Science and Technology of Zhejiang University in collaboration with vivo and other organizations. The framework replaces the traditional U-Net architecture with an innovative Diffusion Transformer (DiT) architecture, combined with a fully self-attentive machine...

Homestyler is the world's top AI home decoration design platform jointly created by Homestyler and Alibaba. Based on AI, 3D, VR and big data technologies, the platform provides convenient online design tools for home decoration designers and DIY customers, and at the same time provides digital...

PandaWiki is an open source AI knowledge base building system that helps users quickly build an intelligent knowledge management system. The system is based on powerful AI big model capabilities, providing AI-assisted creation, Q&A and search functions, significantly improving the efficiency and intelligence of knowledge management.

Anifusion is an online AI manga and anime image creation platform based on AI technology that helps creators produce professional-grade manga and anime artwork without the need for drawing skills. Users simply enter descriptive prompts and AI can automatically generate comic pages and character images. The platform offers intuitive layout...

ZeroSearch is Alibaba Tongyi Lab open source innovative large model search engine framework. The framework does not need to interact with real search engines , based on the simulation of the search engine , with a large model of its own pre-training knowledge to generate relevant or noise documents , significantly reducing the training cost ( reduce 80% or more ...

Shortbread is an AI comic generation tool that allows users to quickly create complete comics based on simple natural language descriptions.Shortbread doesn't require any drawing skills, just type in the text and AI quickly generates personalized comics with unique characters, scenes and emotions.Sho...

ChatUI is an open source intelligent conversational UI component library launched by Alibaba team.ChatUI based on the React framework written in TypeScript , with responsive design , automatically adapted to desktop browsers and mobile devices and other terminals.ChatUI supports internationalization features ...

MotionGo is an AI-driven PPT animation plug-in, based on intelligent dialog and exclusive animation libraries, to help users create presentations quickly and efficiently.MotionGo provides more than 7,000 kinds of animation effects, supports FlowCode function, breaks through the limitations of traditional animation, and realizes the complex motion...

Youdao Little P is an AI all-subject learning assistant launched by NetEase Youdao, designed for K12 students, equipped with the Youdao Ziyi education big model, covering elementary school, middle school and high school all-subject Q&A, providing personalized learning advice. With AI word search and AI translation functions, Youdao Little P helps students quickly solve language problems...

Sparkify is an AI animated video generation platform launched by Google. The platform is based on Gemini 2.5 and Veo2 models, users input questions or complex concepts, Sparkify in 2 minutes to generate intuitive animated short videos to explain the relevant knowledge points. The platform supports text, image and...

AIVA (Artificial Intelligence Virtual Artist) is powerful AI music generation platform that generates over 250 styles of original music in seconds. Users influence the generated music based on uploading audio or MIDI files...

VideoTutor is an AI educational aid designed for K12 students that automatically generates instructional videos with animations and voice-over explanations based on user-entered text, screenshots, or voice-over questions. The tool covers SAT Math, AP Math, STEM knowledge and...

Manga Image Translator is an open source manga image text translation tool that supports one-click translation of the text in a manga or image into the target language.Manga Image Translator can seamlessly embed the original image, maintaining the original style of the manga and cloth...

Index-AniSora is an advanced anime video generation model open source by Beili Beili. The model can generate a coherent animation video based on a single picture , support a variety of styles , such as drama , national animation , VTuber content and so on. The model is based on the diffusion model architecture , combined with the space-time mask module , 3D...

Muzaic is an AI-driven online music generation platform designed for video producers and content creators that enables AI-based technology to quickly generate customized music that matches video content, generating a one-minute studio-quality track in just 8 seconds.

Shear Little Ying is an AI video editing application launched by Byte Jump, designed for zero-basic users, with simple and convenient operation. The application has AI intelligent analysis function, can accurately identify the scene, characters and other elements in the material, to provide users with intelligent editing guidelines. Based on the automated movie function, users can quickly generate high-quality...

Magic Resume is an online AI resume optimization tool based on advanced AI technology to help users quickly create professional and personalized resumes. The tool supports intelligent touch-ups, grammar checking and professional suggestions to automatically enhance the professionalism of textual expressions. All data is stored locally to ensure privacy and security.

Job AI resume is an efficient and convenient intelligent resume generation and optimization platform. Based on AI technology, the platform helps users quickly generate professional and personalized resumes. Users only need to enter basic information and experience, the platform can generate high-quality resume in a short time, providing 2800+ beautiful templates, covering a variety of positions.

NovaCV is an intelligent resume generation tool. The tool is based on advanced AI technology to help users quickly create a professional job resume.AI resume assistant supports one-click generation of job descriptions, intelligent embellishment of the language, optimize the expression of the real-time error correction, provide Chinese and English translation.

V-JEPA 2 is a large model of the world based on video data with 1.2 billion parameters from Meta AI. The model is trained based on self-supervised learning from over 1 million hours of video and 1 million images to understand objects, actions, and movements in the physical world...

Mistral Code is an AI programming assistant for enterprise development teams launched by Mistral AI, integrating the four major models of Codestral, Codestral Embed, Devstral and Mistral Medium, supporting...

imgAK is a one-stop AI image processing tool, based on advanced deep learning technology, providing users with a variety of powerful image editing features. imgAK supports repairing scratches, breaks and fading problems in old photos, and can automatically colorize black and white photos to make them look new. imgAK is a one-stop AI image processing tool, based on advanced deep learning technology, offering users a variety of powerful image editing features.

AopsAI is an online AI tool that transforms old photos into dynamic short videos. Based on advanced AI technology, the tool makes static photos "move", recreating beautiful moments from the past. Users upload photos and provide their email address, and AopsAI automatically generates a video within 24 hours and sends it to...

Genspark AI is an innovative AI browser from Genspark, Inc.Genspark AI comes with a built-in intelligent assistant that helps users find better deals, compare product prices, and analyze user reviews while shopping to help them make smarter purchasing decisions....

Doubao Big Model 1.6 (Doubao-Seed-1.6) is a multimodal deep thinking big model launched by ByteDance. The model supports a variety of input forms such as text, images, videos, etc., and can generate high-quality text output. The model has a 256k long context window, and the maximum input length can be up to 2...

Seedance 1.0 is an advanced video generation model introduced by ByteDance Seed team.Seedance 1.0 supports text and image inputs, generates 1080p high quality videos with seamless multi-camera switching, has native multi-camera narrative capabilities, and realizes far, middle, and near...

AI Pocket Sheet Music is a music processing platform based on AI technology, mainly for music creators, teachers, students and music lovers. The platform supports one-key conversion of audio files into pentatonic scores, rapid separation of vocal and accompaniment tracks, automatic generation of complete musical works based on the melody or chords input by the user, and support for MP3 ...

Splash is an innovative AI music creation platform that helps users easily create personalized music pieces. Based on advanced AI technology, the platform provides text-to-song, text-to-rap and generative text-to-music functions, where users only need to input text and AI can generate melodies and singing effects.

Music Muse is a music creation platform based on advanced AI technology. Users input simple descriptions, such as music style, mood, rhythm, etc., and AI quickly generates music works that meet the needs, without the need for specialized music knowledge. The platform supports a variety of styles such as pop, rock, classical, etc., and can generate music according to the mood of...

Jammable is an AI music cover platform that allows users to upload music or select an existing work and let the AI mimic the voice or style of a favorite artist to create a new cover version.Jammable generates high-quality music quickly, with results in seconds, and features popular sounds, music compilations, and co...

Magistral is an inference model from Mistral AI that focuses on transparent, multilingual and domain-specific reasoning capabilities. The model consists of an open source version (Magistral Small) and an enterprise version (Magistral Medium), the latter in...

o3-pro is an upgraded version of the o3 inference model from OpenAI, designed to handle complex questions and provide accurate answers. The model supports invoking ChatGPT's full suite of tools, such as web search, file analysis, image inference and Python programming, demonstrating powerful execution...

Narakeet is powerful AI text-to-speech platform based on natural speech synthesis technology that supports fast conversion of text, Word documents, PPT presentations or subtitle files to high-quality audio and automatic synchronization with video content.

AIFlowy is an open source enterprise-level AI application development platform, based on Java development, against byte Coze, Tencent meta ware and Dify and other products. Support for intelligent dialog robots, private knowledge base construction, AI workflow orchestration and large model management and other functions, to provide a complete system management model...

Elegant YOYA is a multimodal literate video platform launched by Zhongke Wenge, the platform is based on AI multimodal technology to empower the whole chain of video content creation. Users only need to input the theme requirements, the platform can quickly generate scripts, images, videos, and can complete intelligent editing, voice synthesis and character mouth drive and other operations, the output...

Uthana is a powerful AI 3D character animation generation platform. Users can input text descriptions, upload reference videos, or search motion libraries, and AI can quickly generate realistic animations and support adapting models with any bone structure. The platform is equipped with a variety of features such as style migration, API integration, customization...

Vace AI is a powerful one-stop AI video production and editing platform. The platform generates high-quality video content based on text descriptions or reference images, and supports a variety of editing features such as video expansion, style conversion, subject removal and reconstruction.

Penguin Reading Companion is an AI reading assistant designed for primary and secondary school students by Tencent. Penguin Reading Companion relies on Tencent's hybrid big model and metamachine platform, combined with the Compulsory Education Language Curriculum Program and Curriculum Standards (2022 Edition), to provide students with personalized reading recommendations, multiple reading modes (focusing, reading aloud, listening...

Lemon Slice Live is a real-time video chat AI tool from Lemon Slice, Inc.Lemon Slice Live is based on the Diffusion Transformer Model (DiT), which supports the transformation of any photo or illustration into a real-time interactive animated...

BoomCut is an AI cross-border marketing video localization tool launched by Little Shadow Technology. The tool is equipped with powerful features, based on advanced lip-sync driving and tone cloning technology, to quickly generate multiple language versions of spoken word videos to meet the language needs of users in different countries and regions.

Mobvoi MCP Server is a tool for integrating multimodal AI capabilities launched by GoDoQ. Mobvoi MCP Server is a multimodal AI tool that integrates voice generation, sound cloning, image-driven digital people, video dubbing and video translation, etc., which can be flexibly invoked by users inputting text.Mobvoi MCP S...

Rowboat is an open source low-code AI IDE that supports building multi-intelligent body assistants.Rowboat is based on a visual interface and AI-assisted development features to help users quickly design, configure and test intelligent body workflows. Supporting users to describe requirements in natural language, Row...

Shangtang Ruying is an AI digital human video production platform launched by Shangtang Technology. Based on big model technology, the platform supports the creation of highly realistic digital human images and personalization, including facial features, clothing, hairstyles, and so on. The platform is equipped with sound cloning, video generation, automated data labeling, real-time interaction, and other functions...

JoyHallo is an open source AI digital human model from Jingdong, designed for Mandarin, supporting the conversion of audio into realistic speaking video.JoyHallo embeds audio features based on the wav2vec2 model, using a semi-decoupled structure to improve the accuracy of lip movement prediction, and supports the generation of English video...

Silicon Language AI Knowledge Venture Platform is an AI digital human technology platform launched by Nanjing Silicon Intelligence Technology. The platform relies on advanced AI technologies such as natural semantic processing, machine learning and deep learning to create virtual characters that can interact with humans.

Must Fire AI is a domestic AI digital human generation platform for short video creators. Users can upload a 3-minute video of a real person to quickly generate a highly realistic digital human image, with a micro-expression accuracy of 0.1 millimeters. The platform supports voice synthesis and recording of voice samples to generate AI voice models comparable to real people...

Thetawave AI is an advanced AI note-taking tool designed for college students. Thetawave AI supports real-time capturing of classroom content into structured, easy-to-learn notes, and supports uploading PDF, Word and other documents, which are automatically converted into clear and summarized notes. The main features of the tool include real-time conversion...

Wisdom Spectrum CoCo is the first enterprise-level super assistant Agent launched by Wisdom Spectrum's AICO platform.Wisdom Spectrum CoCo is equipped with three core features: delivery-oriented, memory mechanism and seamless embedding. In the field of government affairs, CoCo can interpret policies, customize solutions and track implementation effects, helping policies to be implemented efficiently.

draw.io is a free online drawing tool that runs on a browser-based platform and does not require any software installation, allowing users to create and edit diagrams anytime, anywhere. draw.io supports a variety of diagram types, such as flowcharts, mind maps, network topology diagrams, Gantt charts, etc., to meet the needs of different scenarios. Operation ...

Forage-Migo is an AI-driven learning and research innovation platform with powerful multimodal AI capabilities supporting text, language, vision and data processing, which can meet the diversified needs across disciplines. Migo provides intelligent Q&A, webpage assisted reading, literature reading, academic writing, thesis reading, chemical...

ScienceOne is an intelligent scientific research platform jointly launched by Institute of Automation, Chinese Academy of Sciences. The platform is based on the construction of large models of scientific foundation, and promotes a new paradigm of intelligent scientific research with multidisciplinary collaboration, providing support for the whole process of scientific research.The core products of ScienceOne include S1...

QBot is a smart browser with integrated AI features launched by Tencent QQ Browser. The browser is equipped with a variety of practical functions, such as AI search, which supports text, voice and image search, and can provide answers quickly and accurately.AI browsing function supports the rapid interpretation of web content to generate mind maps.

FiaMind is an AI mind mapping tool designed for efficient information management and creative assistance. The tool has a flexible unlimited canvas, supports the free splitting and combination of mind maps, to meet the needs of diverse layouts, covering logic diagrams, mind maps, organizational charts and other forms.

Curtain is an AI outline notes tool to help users systematize scattered knowledge points based on a structured outline form, the tool supports one-click generation of mind maps. Curtain support will be complex information to intuitive graphic presentation, help quickly clear thinking, improve understanding and memorization efficiency.

StartAI is an Adobe Photoshop-based AI painting software designed for designers and creative workers. The software is based on AI technology to improve design efficiency and stimulate creativity, and supports text-based drawing, partial redrawing, line coloring, non-destructive zoom, high-definition restoration and other...

Inter-Italian AI is the leading AI drawing and design customization platform that supports bilingual input and contains 400+ models and 100,000 drawing styles. Users input text descriptions or upload images to quickly generate images that meet their requirements.

Metamirror is an AI video creation tool based on the human-computer symbiosis engine, which supports efficient creation from creative inspiration to finished video. The tool is equipped with automated script generation, character style unification, multimodal fusion and intelligent workflow, etc. It can quickly generate creative video scripts, multimodal split-screen design, and synthesize the complete video with one key...

Vermilion Bird AI Detection is an AI detection platform launched by Tencent's hybrid security team, Vermilion Bird Labs, to help users identify AI-generated images and text content. Vermilion Bird AI detection is based on analyzing the hidden features of images, content that does not conform to common sense logic, and "watermark" logos, etc., to quickly determine whether an image is generated by AI.

Mr. Zhejiang University is a platform for deeply integrated intelligences launched by Zhejiang University, based on DeepSeek V3 and R1 models, relying on CARSI resource sharing platform, providing free services for teachers and students of Zhejiang University as well as teachers and students of 829 CARSI-allied universities across the country.

Qin Music Grand Model is an advanced AI music creation grand model jointly launched by Tencent AI Lab and Tencent TME Tianqin Lab. The model intelligently generates high-quality stereo audio or multi-track sheet music based on user-inputted keywords, descriptive statements or audio clips in English and Chinese.

Shoot Me AI is an innovative AI video generation platform launched by Aishi Technology, customized for the domestic market and is the domestic version of PixVerse. The platform supports the rapid generation of high-quality dynamic video content based on simple text prompts or uploaded images. The latest V4.5 version of the platform has improved video quality, animation smoothness...

Often friends ask if there are any tools for generating better quality PPTs. "Making PPT is harder than writing content!"

Nail Yitai is a low-code application development platform launched by Alibaba to help enterprises quickly build digital business applications. Through the visualization of drag and drop and configuration methods, business people who do not know how to code can easily develop applications that meet their needs, greatly reducing the development threshold and cost.

Seed-Music is a big model of AI music generation launched by ByteDance, which supports transforming 10 seconds of user-recorded audio into a complete musical composition. Based on autoregressive language modeling and diffusion methods, it generates multimodal user inputs (e.g., style descriptions, audio references, scores, and sound cues) based on high...

Sound Control is an innovative AI music creation platform that provides comprehensive support for music creators. Audio Control is equipped with various functions such as AI lyrics, compositions, accompaniment generation, professional recording, etc. Users only need to input simple lyrics or melody snippets, and the AI can quickly generate complete song content, covering rock, rap, ballad, and many other...

AntiSpectrum is an innovative online AI music conversion platform, based on advanced AI technology, to convert audio files (such as MP3, FLAC, etc.) into pentatonic and simple scores. AntiSpectrum has a vocal separation function, which separates the vocals from the accompaniment in the music, making it easy for music production and mixing. AntiSpectrum supports converting MIDI files...

Replay is an advanced AI music accompaniment separation tool based on deep learning technology. The tool supports automatically analyzing audio content and accurately separating lead, vocal and backing tracks to generate separate audio files.Replay supports many common audio formats such as MP3, WAV, FL...

MakeBestMusic is an AI-based music creation platform that allows users to quickly generate high-quality instrumental or vocal music based on text descriptions or lyrics.MakeBestMusic supports pop, rock, classical, electronic and other music styles, providing music separation functions...

HunyuanVideo-Avatar is an advanced voice digital human model jointly launched by Tencent Mixed Yuan team and Tencent Music Tianqin Lab. The model is based on the innovative multimodal diffusion Transformer architecture, which generates a natural expression based on the user's uploaded character image and audio...

HeyGen is an AI-driven digital human video creation platform that supports a streamlined video production process, allowing users to quickly generate professional-level digital human videos. The platform is based on advanced AI technology, giving users full control over the image and voice of digital people, providing a rich library of material, including diverse background...

Keevx is a platform for AI digital human video creation, mainly for overseas SMEs and individual creators. Based on AI intelligent script generation and translation functions, with high-quality public portraits and templates, it provides users with one-click digital human marketing video generation services.

Make is an AI-driven no-code automation platform that helps organizations improve efficiency and innovation based on automated processes. The platform offers more than 2,000 pre-built apps that support a variety of business scenarios, such as marketing, sales, finance, etc. Make's core features include no-code visual process creation, AI...

MiMo-VL is Xiaomi's open source multimodal grand model, consisting of a visual coder, a cross-modal projection layer and a language model. The visual coder is based on Qwen2.5-ViT, which supports native resolution inputs and preserves more details; the language model is Xiaomi's self-developed MiMo-7B, which is designed for complex projections...

Olovka AI is an AI academic writing assistance platform for students, which provides accurate writing advice and assistance based on students' academic level, field of specialization and type of paper. Based on intelligent algorithms, Olovka AI helps students quickly write high-quality academic papers that will be...

Fish Audio is a powerful generative AI speech synthesis tool that supports text-to-speech (TTS) and voice cloning. Users only need to input text, the tool supports the conversion to natural and smooth voice, the platform provides multiple languages and voice styles to choose from, to meet different scenarios and user...

SignGemma is the world's most powerful sign language interpreting AI model introduced by Google DeepMind, supporting the accurate translation of American Sign Language (ASL) into English text. The model is based on multimodal training, combining visual and textual data to capture sign language actions in real time and quickly translate them into text...

Jaaz is an open source AI design Agent, local free Lovart pingtai project. With powerful AI design capabilities , can intelligently generate design prompts , batch generate images, posters, storyboards , etc. Jaaz support Ollama, Stable Dif...

CRIC Depth Intelligence is the first AI intelligent body of Chinese real estate independently developed by CRIC, based on CRIC's 20 years of experience in the real estate industry and data accumulation and multimodal big model technology, which opens up the whole chain from data integration, intelligent analysis to content generation.

FLUX.1 Kontext is an image generation and editing model from Black Forest Labs that provides context-aware image processing techniques. The model understands responses to text and image cues, performs tasks such as object modification, style conversion, and background replacement, while maintaining the corner...

WebAgent is an open source autonomous search AI Agent from Alibaba's Tongyi Labs, with powerful end-to-end autonomous information retrieval and multi-step reasoning capabilities.WebAgent can actively perceive, decide and act in the network environment like a human being, and is widely used in academic research, business decision...

Spirit Code IDE is the AI native integrated development environment (IDE) launched by Tongyi Spirit Code, which is deeply adapted to the 3 major models of Thousand Questions, and has a powerful programming intelligent body mode to support the autonomous completion of tasks such as project perception, code retrieval, and execution of terminal operations. It supports MCP tools and integrates Magic Hitch MCP Square's 3...

BAGEL is a multimodal base model open-sourced by ByteDance with 14 billion parameters, of which 7 billion are active. The model base with the Mixed Transformer Expert Architecture (MoT) captures pixel-level and semantic-level features of an image with two independent encoders, respectively, to support efficient processing of images, text, video...

DeepSeek-R1 is a high-performance AI inference model launched by Hangzhou-based DeepSeek, benchmarking against OpenAI's o1 official version. The model is post-trained based on large-scale reinforcement learning techniques and requires only a very small amount of labeled data to reason in math, code and natural language...

Phantom Boat AI is a powerful one-stop AI short film creation platform that supports efficient batch generation of various types of video content, including commercials, promos, animations and more. The platform is based on Midjourney, Runway and other world-leading AI models, and provides creators with a wide range of services from scriptwriting to...

Circuit Tracer is an open source tool from Anthropic for studying the internal workings of large language models. Based on the generation of attribution graphs (attribution graphs) to reveal the internal steps that the model undergoes when generating a particular output ...