anthology-of-the-sota

generative-modeling

Graph View

212 items with this tag.

Oct 03, 2026
LIT-771: Glow: Generative Flow with Invertible 1x1 Convolutions
Oct 03, 2026
LIT-772: Neural Ordinary Differential Equations
Oct 03, 2026
LIT-773: Improving Video Generation with Human Feedback
Oct 03, 2026
LIT-774: Diffusion Model Alignment Using Direct Preference Optimization
- adaptation-and-tuning
- generative-modeling
Oct 03, 2026
LIT-775: Inductive Moment Matching
Oct 03, 2026
LIT-776: Training Diffusion Models with Reinforcement Learning
Oct 03, 2026
LIT-777: Flow Matching Guide and Code
- generative-modeling
- flows-and-transport
Oct 03, 2026
LIT-778: DPOK: Reinforcement Learning for Fine-tuning Text-to-Image Diffusion Models
- adaptation-and-tuning
- generative-modeling
Oct 03, 2026
LIT-779: Flow-GRPO: Training Flow Matching Models via Online RL
Oct 03, 2026
LIT-780: FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Models
- generative-modeling
- flows-and-transport
Oct 03, 2026
LIT-781: Density estimation using Real NVP
Oct 03, 2026
LIT-782: Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generation
Oct 03, 2026
LIT-783: Large Scale Diffusion Distillation via Score-Regularized Continuous-Time Consistency
Oct 03, 2026
LIT-784: Mean Flows for One-step Generative Modeling
Oct 03, 2026
LIT-785: STARFlow: Scaling Latent Normalizing Flows for High-resolution Image Synthesis
Oct 03, 2026
LIT-786: Improved Techniques for Training Consistency Models
Oct 03, 2026
LIT-787: One Step Diffusion via Shortcut Models
Oct 03, 2026
LIT-788: ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation
Oct 03, 2026
LIT-789: Multisample Flow Matching: Straightening Flows with Minibatch Couplings
Oct 03, 2026
LIT-790: Simplifying, Stabilizing and Scaling Continuous-Time Consistency Models
Oct 03, 2026
LIT-791: Improving the Training of Rectified Flows
Oct 03, 2026
LIT-792: Discrete Flow Matching
Oct 03, 2026
LIT-793: Flow map matching with stochastic interpolants: A mathematical framework for consistency models
Oct 03, 2026
LIT-794: Normalizing Flows are Capable Generative Models
Oct 03, 2026
LIT-796: InstaFlow: One Step is Enough for High-Quality Diffusion-Based Text-to-Image Generation
Oct 03, 2026
LIT-797: Aligning Text-to-Image Models using Human Feedback
Oct 03, 2026
LIT-798: Improving and generalizing flow-based generative models with minibatch optimal transport
Oct 03, 2026
SOTA-442: To RL-tune a flow-matching generator, roll it out as a same-marginal SDE built from its own velocity, and anchor each step with the closed-form KL to the reference
Oct 03, 2026
SOTA-443: Down-weight large residuals in consistency and MeanFlow training: use a Pseudo-Huber loss or an adaptive inverse-error weight, not plain squared L2
Oct 03, 2026
SOTA-444: When a flow will be sampled with a coarse ODE solver, pair noise and data within each minibatch by exact optimal transport
Oct 03, 2026
SOTA-445: When the base text-to-image diffusion model is better than the models that generated the preference pairs, train on the pairs with Diffusion-DPO rather than fine-tuning on the preferred images
- adaptation-and-tuning
- generative-modeling
Oct 03, 2026
SOTA-446: In consistency training, take the target from the current weights under stop-gradient, not from an EMA teacher
Oct 03, 2026
SOTA-447: When distilling a large image or video model by continuous-time consistency, add a small DMD distribution-matching term
Oct 03, 2026
SOTA-448: Without a pretrained teacher, train a one-step generator as a MeanFlow average-velocity model rather than by consistency training or a shortcut model
Oct 03, 2026
SOTA-449: When reward-tuning a classifier-free-guided diffusion model over several rounds of sampling, train the guided prediction at the fixed sampling guidance weight, not the conditional prediction alone
- adaptation-and-tuning
- generative-modeling
Oct 03, 2026
THEORY-118: Continuous-time consistency training is unstable through the time derivative in its tangent, and tangent normalization, which MeanFlow's adaptive loss weight already is, controls it
Oct 03, 2026
THEORY-119: Consistency, shortcut, MeanFlow and progressive-distillation models all estimate the two-time flow map of the probability-flow ODE, and differ in which identity of that map they train on
Oct 03, 2026
THEORY-120: A likelihood-trained normalizing flow samples well only when its training data carry Gaussian noise well above the quantization width, because its inverse must be well conditioned over the whole Gaussian latent
- generative-modeling
- flows-and-transport
Oct 03, 2026
THEORY-121: Distribution-matching distillation loses coverage because it descends a reverse divergence on the student's own samples, and a consistency term on teacher trajectories restores it
- generative-modeling
- few-step-generation
Oct 03, 2026
THEORY-122: A diffusion or flow sampler can be trained by a likelihood-ratio policy gradient only when each step is stochastic with a tractable density, and a same-marginal SDE built from the trained velocity supplies that without changing the marginals
Oct 03, 2026
THEORY-123: A flow's sampling ODE curves because the noise-data pairs it was trained on cross and the learned velocity averages them, so a coupling whose straight paths do not cross gives a straight flow
Oct 03, 2026
THEORY-124: Consistency training is a bootstrapped fixed-point iteration, not descent on a loss, so a target that lags the student swamps a data signal that enters only at the order of the step
Oct 03, 2026
THEORY-125: One affine autoregressive flow block makes every conditional a single Gaussian and cannot approximate an arbitrary density, while two blocks in opposite orders make every conditional but the last a Gaussian mixture and three make all of them one
Oct 02, 2026
LIT-770: DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generation
Oct 01, 2026
LIT-743: Lyra: Generative 3D Scene Reconstruction via Video Diffusion Model Self-Distillation
Oct 01, 2026
LIT-751: "Principal Components" Enable A New Language of Images
Oct 01, 2026
LIT-755: Wonderland: Navigating 3D Scenes from a Single Image
Oct 01, 2026
SOTA-440: Supervise a feed-forward 3D reconstructor with a frozen camera-controlled video model's renderings instead of captured multi-view data
Sep 26, 2026
LIT-722: Score-Based Generative Modeling through Stochastic Differential Equations
Sep 25, 2026
LIT-676: DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models
Sep 25, 2026
LIT-677: Diffusion Models With Learned Adaptive Noise
- generative-modeling
- training-optimization
Sep 25, 2026
LIT-678: Diffusion Meets Flow Matching: Two Sides of the Same Coin
Sep 25, 2026
LIT-687: Fine-Tuning Image-Conditional Diffusion Models is Easier than You Think
Sep 25, 2026
LIT-689: Common Diffusion Noise Schedules and Sample Steps are Flawed
Sep 25, 2026
LIT-691: Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation
Sep 25, 2026
LIT-692: Understanding Diffusion Objectives as the ELBO with Simple Data Augmentation
Sep 25, 2026
LIT-693: Classifier-Free Diffusion Guidance
Sep 25, 2026
LIT-699: Diffusion Models Beat GANs on Image Synthesis
Sep 25, 2026
LIT-701: Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution
Sep 25, 2026
LIT-702: Simple and Effective Masked Diffusion Language Models
Sep 25, 2026
LIT-703: Improved Precision and Recall Metric for Assessing Generative Models
Sep 25, 2026
LIT-710: Efficient Diffusion Training via Min-SNR Weighting Strategy
- generative-modeling
- training-optimization
Sep 25, 2026
LIT-714: Analyzing and Improving the Training Dynamics of Diffusion Models
Sep 25, 2026
LIT-720: Analyzing and Improving the Training Dynamics of Diffusion Models
Sep 25, 2026
LIT-721: Guiding a Diffusion Model with a Bad Version of Itself
Sep 25, 2026
SOTA-410: For guided sampling, use a second-order multistep solver on the data-prediction parameterization, not a higher-order solver on the noise prediction
- inference-optimization
- generative-modeling
Sep 25, 2026
SOTA-412: When the target is likelihood, learn a per-dimension noise schedule conditioned on a learned latent; a scalar or unconditioned one buys nothing
- generative-modeling
- training-optimization
Sep 25, 2026
SOTA-416: When sampling a diffusion model in few steps, start at t = T: use trailing, not leading, timestep spacing
Sep 25, 2026
SOTA-422: Rescale the noise schedule so the last timestep is pure noise, and train with v-prediction so the model can learn there
- generative-modeling
- training-optimization
Sep 25, 2026
SOTA-423: When classifier-free guidance over-exposes a latent diffusion model, rescale the guided prediction toward the conditional prediction's standard deviation
- generative-modeling
- inference-optimization
Sep 25, 2026
SOTA-424: Train one network for both conditional and unconditional scores by dropping the condition on 10% of examples, then pick the guidance weight by which metric you are willing to lose
- generative-modeling
- training-optimization
Sep 25, 2026
SOTA-425: Report precision and recall beside FID whenever the generator has a fidelity-diversity knob, because FID's best value sits in the interior of that trade
- analysis-and-evaluation
- generative-modeling
Sep 25, 2026
SOTA-428: Choose the EMA length after training: store two power-function averages at snapshots and reconstruct any length by least squares
Sep 25, 2026
SOTA-431: Make every layer of a diffusion denoiser preserve activation magnitude, and force each weight vector back to fixed norm after every step
Sep 25, 2026
SOTA-432: Maintain two power-function weight averages during training and snapshot them, so the averaging length is a post-hoc sweep instead of a guess you have to make before the run
Sep 25, 2026
SOTA-433: Guide a diffusion model with a smaller, less-trained copy of itself rather than an unconditional model, and make sure the two models are degraded in the same way
Sep 25, 2026
THEORY-104: A large guidance scale amplifies the model's derivatives, which narrows a high-order solver's convergence radius
Sep 25, 2026
THEORY-106: Gaussian flow matching is a diffusion model; what separates the named recipes is loss weighting, network output and sampling schedule
Sep 25, 2026
THEORY-107: A diffusion loss whose weighting falls monotonically with signal-to-noise is maximum likelihood on noise-augmented data; a non-monotone one is not a likelihood objective
- generative-modeling
- training-optimization
Sep 25, 2026
THEORY-109: Classifier guidance may flatter classifier-based metrics because it steps along a classifier gradient, and guidance without a classifier is the control
- analysis-and-evaluation
- generative-modeling
Sep 25, 2026
THEORY-110: FID is not neutral between fidelity and coverage: it weights coverage more, so optimizing it walks toward variety at the cost of per-sample quality
- analysis-and-evaluation
- generative-modeling
Sep 25, 2026
THEORY-112: Classifier-free guidance improves image quality because its unconditional reference model is worse, not because it emphasizes the class, so guidance is adaptive truncation toward the better-fit density
- generative-modeling
- analysis-and-evaluation
Sep 24, 2026
LIT-618: LTX-Video: Realtime Video Latent Diffusion
Sep 24, 2026
LIT-619: Wan: Open and Advanced Large-Scale Video Generative Models
Sep 24, 2026
LIT-620: HunyuanVideo: A Systematic Framework For Large Video Generative Models
Sep 24, 2026
LIT-621: Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Models
Sep 24, 2026
LIT-622: CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer
Sep 24, 2026
LIT-623: Language Model Beats Diffusion — Tokenizer is Key to Visual Generation
Sep 24, 2026
LIT-624: Step-Video-T2V Technical Report: The Practice, Challenges, and Future of Video Foundation Model
Sep 24, 2026
LIT-625: Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets
Sep 24, 2026
LIT-626: Movie Gen: A Cast of Media Foundation Models
Sep 24, 2026
LIT-627: Video Diffusion Models
- generative-modeling
- vision-and-graphics
Sep 24, 2026
LIT-629: Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion
Sep 24, 2026
LIT-630: Flow Matching for Generative Modeling
- generative-modeling
- flows-and-transport
Sep 24, 2026
LIT-631: From Slow Bidirectional to Fast Autoregressive Video Diffusion Models
Sep 24, 2026
LIT-632: Make-A-Video: Text-to-Video Generation without Text-Video Data
Sep 24, 2026
LIT-633: AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models without Specific Tuning
Sep 24, 2026
LIT-634: Open-Sora 2.0: Training a Commercial-Level Video Generation Model in $200k
Sep 24, 2026
LIT-635: Emu Video: Factorizing Text-to-Video Generation by Explicit Image Conditioning
- generative-modeling
- vision-and-graphics
Sep 24, 2026
LIT-636: Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow
Sep 24, 2026
LIT-637: Imagen Video: High Definition Video Generation with Diffusion Models
Sep 24, 2026
LIT-642: Professor Forcing: A New Algorithm for Training Recurrent Networks
- training-optimization
- generative-modeling
Sep 24, 2026
LIT-643: One-step Diffusion with Distribution Matching Distillation
Sep 24, 2026
LIT-644: Building Normalizing Flows with Stochastic Interpolants
Sep 24, 2026
LIT-645: Stochastic Interpolants: A Unifying Framework for Flows and Diffusions
Sep 24, 2026
LIT-646: Improved Distribution Matching Distillation for Fast Image Synthesis
Sep 24, 2026
LIT-647: Sequence Level Training with Recurrent Neural Networks
Sep 24, 2026
LIT-648: Improving Image Generation with Better Captions
Sep 24, 2026
LIT-649: Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks
- training-optimization
- generative-modeling
Sep 24, 2026
LIT-652: Video generation models as world simulators
Sep 24, 2026
LIT-658: FiT: Flexible Vision Transformer for Diffusion Model
Sep 24, 2026
LIT-659: On the Importance of Noise Scheduling for Diffusion Models
- generative-modeling
- signal-structure
Sep 24, 2026
LIT-660: simple diffusion: End-to-end diffusion for high resolution images
Sep 24, 2026
LIT-661: Photorealistic Video Generation with Diffusion Models
Sep 24, 2026
SOTA-386: Show a video diffusion model images before and alongside video
Sep 24, 2026
SOTA-389: Caption training video with a model that watches the video, not with captions of its frames
Sep 24, 2026
SOTA-390: In a video diffusion transformer, attend over space and time jointly rather than factorizing, and budget for the cost
Sep 24, 2026
SOTA-392: To get a one-step sampler from a rectified flow, distil it; treat reflow as an optional extra pass, and keep the pre-reflow model for many steps
Sep 24, 2026
SOTA-393: Freeze the image model's spatial layers when adding time if checkpoint compatibility matters; do not expect a quality gain from it
Sep 24, 2026
SOTA-394: When distilling a causal few-step video generator, take the teacher bidirectional, not causal
Sep 24, 2026
SOTA-395: Post-train an autoregressive video generator on its own rollouts, and score the whole rollout
Sep 24, 2026
SOTA-396: When a latent video model reuses an image autoencoder or upsampler, fine-tune them on video so they see time
Sep 24, 2026
SOTA-397: To condition an unconditionally trained diffusion model on known frames, use reconstruction guidance, not replacement
Sep 23, 2026
LIT-553: Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise
- generative-modeling
- analysis-and-evaluation
Sep 23, 2026
LIT-554: Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion
- generative-modeling
- vision-and-graphics
Sep 23, 2026
LIT-555: Warm Diffusion: Recipe for Blur-Noise Mixture Diffusion Models
Sep 23, 2026
LIT-559: Alias-Free Generative Adversarial Networks
Sep 23, 2026
LIT-560: Analyzing and Improving the Image Quality of StyleGAN
Sep 23, 2026
LIT-561: A Style-Based Generator Architecture for Generative Adversarial Networks
Sep 23, 2026
LIT-562: Projected GANs Converge Faster
Sep 23, 2026
LIT-563: The Role of ImageNet Classes in Fréchet Inception Distance
- analysis-and-evaluation
- generative-modeling
Sep 23, 2026
LIT-564: Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks
- generative-modeling
- vision-and-graphics
Sep 23, 2026
LIT-565: FLUX-Reason-6M & PRISM-Bench: A Million-Scale Text-to-Image Reasoning Dataset and Comprehensive Benchmark
Sep 23, 2026
LIT-566: SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis
- generative-modeling
- vision-and-graphics
Sep 23, 2026
LIT-572: FLUX.2: Analyzing and Enhancing the Latent Space of FLUX – Representation Comparison
Sep 23, 2026
LIT-573: FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space
Sep 23, 2026
LIT-584: Accurate structure prediction of biomolecular interactions with AlphaFold 3
Sep 23, 2026
LIT-592: Latent Dirichlet Allocation
Sep 23, 2026
LIT-611: GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
Sep 23, 2026
LIT-614: Improved Techniques for Training GANs
Sep 23, 2026
LIT-615: Demystifying MMD GANs
Sep 23, 2026
SOTA-333: For autoregressive generation of continuous sequences, train with an independent noise level per token and condition the rollout on slightly noised history
- generative-modeling
- vision-and-graphics
Sep 23, 2026
SOTA-334: In a style-modulated generator, demodulate the convolution weights instead of instance-normalizing the activations
- generative-modeling
- model-architecture
Sep 23, 2026
SOTA-335: When generated content must move continuously, as in video or animation, build the generator alias-free: treat features as continuous signals and low-pass filter around every nonlinearity
Sep 23, 2026
SOTA-336: Train high-resolution GANs at a fixed topology with output skips in the generator and a residual discriminator, not by progressive growing
Sep 23, 2026
SOTA-337: When an ImageNet classifier is anywhere in the pipeline — training the generator or steering its sampling — confirm FID gains with a Fréchet distance in a non-ImageNet feature space
- analysis-and-evaluation
- generative-modeling
Sep 23, 2026
SOTA-338: Train the GAN discriminator on frozen multi-scale pretrained features with fixed random channel and scale mixing
Sep 23, 2026
SOTA-339: For unpaired image-to-image translation of appearance, constrain both mappings with a cycle-consistency loss
- generative-modeling
- vision-and-graphics
Sep 23, 2026
SOTA-341: Condition an image generator on each training image's original size instead of discarding or upsampling small images
Sep 23, 2026
SOTA-346: Retune the training timestep shift for each autoencoder latent space, and never rank latent spaces for generation under one shared schedule
Sep 23, 2026
SOTA-352: When a structure predictor is made generative, add a regression model's predictions to the training data so disordered regions are learned as disorder rather than invented structure
Sep 23, 2026
SOTA-383: Fix the sample count before comparing FID values, and settle a close comparison with an unbiased estimator rather than a tighter error bar
Sep 23, 2026
THEORY-079: What makes a diffusion model generative is iterated restoration of any degradation, not Gaussian noise
- generative-modeling
- analysis-and-evaluation
Sep 23, 2026
THEORY-080: A diffusion process needs noise to keep its intermediate states on the data manifold, and blur helps only while noise still dominates the bands it removes
Sep 23, 2026
THEORY-086: A text document is better modelled as a mixture of several topics, each a distribution over words, than as a draw from a single topic
- signal-structure
- generative-modeling
Sep 23, 2026
THEORY-095: No unbiased estimator of FID exists, so its plug-in bias can reverse a comparison while the sample variance stays small
Sep 21, 2026
LIT-477: NoProp: Training Neural Networks without Full Back-propagation or Full Forward-propagation
Sep 21, 2026
LIT-479: The Diffusion Duality
Sep 21, 2026
LIT-483: Mixture-of-Transformers: A Sparse and Scalable Architecture for Multi-Modal Foundation Models
Sep 21, 2026
LIT-488: gen2seg: Generative Models Enable Generalizable Instance Segmentation
Sep 21, 2026
LIT-490: Denoising as Projection: Constrained Optimization with Gradient-Guided Diffusion
- generative-modeling
- analysis-and-evaluation
Sep 21, 2026
LIT-491: Noise Hypernetworks: Amortizing Test-Time Compute in Diffusion Models
Sep 21, 2026
LIT-494: GaussianToken: An Effective Image Tokenizer with 2D Gaussian Splatting
Sep 21, 2026
LIT-496: Taming Transformers for High-Resolution Image Synthesis
Sep 21, 2026
LIT-497: Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation
Sep 21, 2026
LIT-498: MaskGIT: Masked Generative Image Transformer
Sep 21, 2026
LIT-499: Neural Discrete Representation Learning
- representation-and-encoding
- generative-modeling
Sep 21, 2026
LIT-500: Vector-quantized Image Modeling with Improved VQGAN
Sep 21, 2026
LIT-501: The FID Lottery: Quantifying Hidden Randomness in Generative-Model Evaluation
Sep 21, 2026
SOTA-289: When the sampling budget is small, prefer uniform-state discrete diffusion with consistency distillation — masked diffusion cannot revise what it has already emitted
Sep 21, 2026
SOTA-301: Apply the objective gradient before the denoiser, not after it, when guiding diffusion toward a task objective
- generative-modeling
Sep 21, 2026
SOTA-302: Steer a distilled generator by modulating its input noise, not by fine-tuning its weights
Sep 21, 2026
SOTA-306: Factorize the codebook: look up in a low-dimensional normalized space, embed in a high-dimensional one, and report utilization alongside reconstruction quality
- representation-and-encoding
- generative-modeling
Sep 21, 2026
SOTA-307: Report generative FID as an error bar over several training seeds, and treat any gap below about 2% of the mean as inconclusive
- analysis-and-evaluation
- generative-modeling
Sep 21, 2026
THEORY-046: Uniform-state discrete diffusion is the argmax of a Gaussian diffusion, which is why Gaussian technique transfers to it and not to masked diffusion
- generative-modeling
- model-architecture
Sep 21, 2026
THEORY-053: The Stein posterior-mean denoiser acts as an approximate projection onto the learned data geometry
- generative-modeling
- analysis-and-evaluation
Sep 21, 2026
THEORY-054: The regularizer that is intractable in data space is tractable in noise space, and bounds the data-space divergence
- generative-modeling
- analysis-and-evaluation
Sep 20, 2026
LIT-446: Variational Diffusion Models
- generative-modeling
Sep 20, 2026
LIT-447: SiT: Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformers
- generative-modeling
- flows-and-transport
Sep 20, 2026
LIT-448: Scalable Diffusion Models with Transformers
- generative-modeling
- model-architecture
Sep 20, 2026
LIT-449: Scaling Rectified Flow Transformers for High-Resolution Image Synthesis
Sep 20, 2026
SOTA-262: Give each modality its own weights and let the streams attend jointly
Sep 20, 2026
SOTA-263: Shift the timestep schedule when the resolution changes, because more pixels need more noise
Sep 20, 2026
SOTA-264: Fix the noise schedule's endpoints against the bound and choose its shape to minimise the loss estimator's variance
- generative-modeling
Sep 20, 2026
SOTA-265: Tune the stochastic sampler's diffusion coefficient after training; it is not fixed by the forward process
- generative-modeling
- flows-and-transport
Sep 20, 2026
SOTA-266: Connect data and noise on a straight line, and sample the training timesteps from a logit-normal rather than uniformly
- generative-modeling
- flows-and-transport
Sep 20, 2026
THEORY-027: In continuous time the diffusion bound depends on the noise schedule only through its endpoints
- generative-modeling
Sep 19, 2026
LIT-439: Deep Unsupervised Learning using Nonequilibrium Thermodynamics
- generative-modeling
Sep 19, 2026
LIT-442: Diffusion Beats Autoregressive in Data-Constrained Settings
- generative-modeling
- data-pipeline
Sep 19, 2026
SOTA-254: Train the language model as a masked diffusion model when the corpus is fixed and the compute is not
Sep 17, 2026
LIT-381: Scaling Diffusion Language Models via Adaptation from Autoregressive Models
- generative-modeling
- adaptation-and-tuning
Sep 17, 2026
SOTA-232: Convert an autoregressive model into a diffusion model by continual pretraining rather than training one from scratch
- generative-modeling
Sep 10, 2026
SOTA-195: Predict v rather than the noise when the model will be evaluated at low signal-to-noise
- training-optimization
- generative-modeling
Sep 10, 2026
SOTA-202: Clamp the prediction to the training range at every step when sampling from a model's own output
- training-optimization
- generative-modeling
Sep 10, 2026
SOTA-203: Sample a diffusion model with a higher-order ODE solver on the weights you already trained
- generative-modeling
Sep 10, 2026
SOTA-204: Anneal a discretisation from coarse to fine over training rather than fixing it
- training-optimization
- generative-modeling
Sep 10, 2026
SOTA-206: Keep a multi-step sampling option in a few-step generative model
- generative-modeling
- few-step-generation
Sep 10, 2026
SOTA-207: Use the deterministic sampler when the noise input has to mean something
- generative-modeling
Sep 08, 2026
SOTA-187: Train the generative model in a learned compressed latent, not at full resolution
Sep 08, 2026
SOTA-188: Parametrize the network so its prediction target has unit variance at every noise level, and sample training noise from a log-normal
- training-optimization
- generative-modeling
Sep 07, 2026
LIT-217: Large Language Diffusion Models
- generative-modeling
Sep 07, 2026
SOTA-157: Train the language model as a masked diffusion model rather than autoregressively
- model-architecture
- generative-modeling
Aug 24, 2026
LIT-036: Denoising Diffusion Probabilistic Models
- generative-modeling
Aug 24, 2026
LIT-038: Denoising Diffusion Implicit Models
- generative-modeling
Aug 24, 2026
LIT-062: High-Resolution Image Synthesis with Latent Diffusion Models
- generative-modeling
- signal-structure
Aug 24, 2026
LIT-067: Progressive Distillation for Fast Sampling of Diffusion Models
- generative-modeling
- few-step-generation
Aug 24, 2026
LIT-070: Hierarchical Text-Conditional Image Generation with CLIP Latents
- generative-modeling
Aug 24, 2026
LIT-073: Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding
- generative-modeling
Aug 24, 2026
LIT-075: Elucidating the Design Space of Diffusion-Based Generative Models
- generative-modeling
Aug 24, 2026
LIT-076: DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps
- generative-modeling
- few-step-generation
Aug 24, 2026
LIT-093: Consistency Models
- generative-modeling
- few-step-generation
Aug 24, 2026
LIT-097: One Transformer Fits All Distributions in Multi-Modal Diffusion at Scale
- generative-modeling

Created with Quartz v5.0.0 © 2026

Repository
Quartz