Diffusion Model
A diffusion model is a probabilistic generative machine learning model architecture whose core principle involves gradually adding noise to data (forward process) and then learning to progressively denoise and reconstruct the data (reverse process) to model the data distribution. Originally applied extensively to image generation, diffusion models have since expanded to generative tasks across modalities including audio, video, and text, and are characterized by high generation quality and strong diversity.
Core Facts
Reasoning models perform better in information synthesis tasks requiring multi-step logical deduction, but have higher latency and should be selected appropriately based on task characteristics.
Diffusion models circumvent the instability of adversarial training through a Markov chain mechanism that progressively adds Gaussian noise and then trains the model to learn the reverse denoising process.
扩散模型的核心原理分为前向扩散过程（逐步添加高斯噪声）和反向去噪过程（从噪声中还原图像）两个阶段
Timeline (last 90 days)
穿模的根源在于扩散或自回归生成模型优化像素/帧级别似然，而非显式刚体约束，长时间滚动后误差会累积放大。
Changing the diffusion model's decoding rule to prioritize filling the cells it is most confident about can raise its Sudoku accuracy from about 25% to close to NCM's 86%, and with more iterations it can reach 95%.
In the paper, NCM's Sudoku advantage mainly comes from the cell-filling order (the decoding strategy) rather than the architecture itself.
"Generative artificial intelligence" refers to a class of models capable of generating text, images, audio and other content based on prompts, represented by GPT and diffusion models.
Mainstream image/video generation is mostly based on diffusion processes and requires multiple denoising iterations, with a single generation often taking from several seconds to tens of seconds; even with distillation-based acceleration, low latency in continuous interactive scenarios remains a challenge.
Real-time interactive world models typically need to adopt streaming inference schemes based on autoregressive or diffusion models, combined with state compression (such as encoding historical frames into compact latent representations) to reduce computational burden.
Diffusion models and score-based methods are particularly attractive for turbulence because they can represent multimodal distributions.
Through recursion, the direct relationship between Xt and X0 at any time step can be derived: X_t = √(α̅_t) * X0 + √(1-α̅_t) * Z, allowing Xt to be sampled without iteration.
The original diffusion model paper used UNet as the noise prediction network and incorporated attention mechanisms.
In diffusion models, the input to UNet is the noisy image X_t, and the output is a predicted noise map of the same size as the input; the time step t is embedded via sinusoidal positional encoding and injected into each residual block.
40 more timeline events
All Facts (20)
Reasoning models perform better in information synthesis tasks requiring multi-step logical deduction, but have higher latency and should be selected appropriately based on task characteristics.
90%VerifiedDiffusion models circumvent the instability of adversarial training through a Markov chain mechanism that progressively adds Gaussian noise and then trains the model to learn the reverse denoising process.
90%Verified扩散模型的核心原理分为前向扩散过程（逐步添加高斯噪声）和反向去噪过程（从噪声中还原图像）两个阶段
90%VerifiedDDPM (Denoising Diffusion Probabilistic Models) laid the theoretical foundation for diffusion models in 2020.
90%VerifiedStable Diffusion is based on the Latent Diffusion Model and generates high-quality images by performing a denoising process in a compressed latent space
90%VerifiedCompared to GANs, diffusion models are more stable to train, generate greater diversity, and are less prone to mode collapse.
90%VerifiedFlow Matching continuously maps a noise distribution to a data distribution by learning a deterministic velocity field, and both Stable Diffusion 3 and Flux adopt this framework.
85%VerifiedDiffusion models are inspired by non-equilibrium thermodynamics and employ a two-step training process: the forward process gradually adds Gaussian noise to an image until it becomes pure random noise, and the reverse process trains a neural network to progressively denoise from pure noise back to a clear image.
80%VerifiedCompared to GANs, diffusion models offer more stable training, higher generation diversity, and easier integration with text conditioning.
80%VerifiedQwen's image generation capability is based on Diffusion Model technology.
80%VerifiedText instructions are converted into semantic vectors through cross-modal encoders such as CLIP, embedding text and images into the same semantic space to guide image generation.
80%VerifiedGANs improve generation quality through adversarial interplay between a generator and a discriminator, but are prone to mode collapse; diffusion models have more stable training objectives and naturally avoid mode collapse.
80%VerifiedThe core idea of diffusion models originates from non-equilibrium thermodynamics: during the training phase, Gaussian noise is gradually added to clear images, and during the inference phase, noise is progressively removed from random noise in reverse to restore the image.
80%VerifiedCurrent mainstream text-to-video models are generally based on diffusion model architectures, generating coherent video frame sequences by extending image generation capability along the temporal dimension.
80%VerifiedThe output randomness of image generation models fundamentally stems from the initial noise sampling in the diffusion process, with the Seed (random number seed) controlling the initial noise.
80%VerifiedThe theoretical foundations of diffusion models can be traced back to a 2015 paper on non-equilibrium thermodynamics published by Sohl-Dickstein et al.
80%VerifiedThe denoising process of diffusion models is based on the DDPM principle, which involves progressively adding Gaussian noise to data during training and then training a neural network to learn the reverse denoising process.
80%VerifiedThe results of AI image expansion differ each time, and this randomness stems from the diffusion model starting from different random noise each time
80%VerifiedMost mainstream AI image generation models are currently based on the Diffusion Model architecture.
80%VerifiedAI video generation technology has undergone a paradigm shift from GANs (Generative Adversarial Networks) to Diffusion Models.
80%