NeurIPS 2026 Papers

Discrete diffusion models are typically trained by predicting the clean data given a noisy observation, which is then used with a certain choice of reverse transition parameterization. Two choices coexist in the literature, which we refer to as marginalization and plug-in. Although they coincide in the case of masked diffusion models (MDMs), we revisit this question for uniform diffusion models (UDMs), where the distinction matters. We show that the plug-in parameterization is not optimized by the denoising posterior but by a \emph{leave-one-out} posterior that removes the influence of the local observation. We give a thorough characterization of the resulting training target, together with exact conversion formulas between the denoiser, the \emph{leave-one-out} posterior, and the score. As a by-product, the leave-one-out characterization yields a Gibbs-based predictor–corrector scheme at no additional training cost. We further propose an absorbing-state construction of uniform diffusion, whose denoiser decomposes into simpler absorbing denoising posteriors, and which recovers the full UDM joint law following a simple resampling step. Empirically, we show that these design choices interact strongly with optimization and generative quality; on language modeling, the \emph{leave-one-out} denoiser improves over other parameterizations.

Flexible molecular docking is crucial for drug discovery; however, existing flow-matching generative models often produce physically invalid conformations (e.g., steric clashes, distorted bond geometry). Our analysis suggests that a primary contributor to this challenge lies at the modeling level: when the flow is parameterized by standard equivariant networks, an intrinsic approximation floor can emerge, potentially hindering the learned velocity field from reaching its optimal convergence bound, particularly for highly flexible molecules. We present UniFlowDock, which addresses this by introducing strictly Complete Equivariant Velocity Fields, thereby eliminating the dimensional incompleteness of standard architectures. This theoretical completeness is further bolstered by Fragment-Adaptive Optimal Coupling, a method that structurally linearizes transport trajectories and decomposes the generation process into physics-aware subproblems. Experiments on PDBBind and PoseBusters demonstrate UniFlowDock achieves state-of-the-art accuracy and physical validity, showing consistent gains in highly flexible docking scenarios where target distributions exhibit higher transport complexity.

Multi-reference image generation aims to synthesize images from textual instructions while faithfully preserving subject identities from multiple reference images. Existing VLM-enhanced diffusion models commonly rely on decoupled visual conditioning: semantic ViT features are processed by the VLM for instruction understanding, whereas appearance-rich VAE features are injected later into the diffusion backbone. Despite its intuitive design, this separation makes it difficult for the model to associate each semantically grounded subject with visual details from the correct reference image. As a result, the model may recognize which subject is being referred to, but fail to preserve its identity and fine-grained appearance, leading to attribute leakage and cross-reference confusion in complex multi-reference settings. To address this issue, we propose UniCustom, a unified visual conditioning framework that fuses ViT and VAE features before VLM encoding. This early fusion exposes the VLM to both semantic cues and appearance-rich details, enabling its hidden states to jointly encode the referred subject and corresponding visual appearance with only a lightweight linear fusion layer. To learn such unified representations, we adopt a two-stage training strategy: reconstruction-oriented pretraining that preserves reference-specific appearance details in the fused hidden states, followed by supervised finetuning on single- and multi-reference generation tasks. We further introduce a slot-wise binding regularization that encourages each image slot to preserve low-level details of its corresponding reference, thereby reducing cross-reference entanglement. Experiments on two multi-reference generation benchmarks demonstrate that UniCustom consistently improves subject consistency, instruction following, and compositional fidelity over strong baselines. Our code, checkpoints, and the training dataset will be released soon.

This position paper argues that understanding generalization in diffusion models requires fundamentally new theoretical frameworks that go beyond both classical statistical learning theory and the benign overfitting paradigm developed for supervised learning. In diffusion models, unlike in supervised learning, memorization of training data and generalization to novel samples are incompatible: a model that has fully memorized its training set generates copies rather than novel data. Several theoretical explanations for why practical diffusion models nevertheless generalize have been proposed, based on capacity limitations, implicit regularization from optimization, or architectural inductive biases, but their interactions remain unclear. We argue that the field should pivot from explaining why the diffusion models do not memorize to investigating what the models actually learn during pre-memorization phase. To highligh our stance, we conduct empirical study of diffusion models trained on CIFAR-10, and we distill the findings into concrete open questions that we believe are key to improve understanding of generalization in diffusion models.

Large Language Diffusion Models (LLDMs) are emerging as an alternative to autoregressive models, offering faster inference through higher parallelism. Similar to autoregressive LLMs, they remain prone to hallucinations, making reliable uncertainty quantification (UQ) crucial for safe deployment. However, existing UQ methods are fundamentally misaligned with this new paradigm: they assume autoregressive factorization or use expensive multiple sampling, negating the efficiency of LLDMs. In this work, we present the first systematic study of UQ for LLDMs and propose lightweight, zero-shot uncertainty signals derived from the iterative denoising process, leveraging intermediate generations, token remasking dynamics, and denoising complexity. We further adapt a state-of-the-art UQ method to LLDMs by combining masked diffusion likelihoods with trajectory-based semantic dissimilarity. We provide theoretical grounding, proving that expected trajectory dissimilarity bounds the masked diffusion training objective. Comprehensive experiments across three tasks, eight datasets, and two models show that our method achieves a great cost-performance trade-off: it approaches the strongest sampling-based baselines while incurring up to one hundred times lower computational overhead. Our work demonstrates that LLDMs can deliver both fast inference and reliable hallucination detection simultaneously.

This work presents UGD-FM, a novel unaligned image guided denoising framework based on cross-modal conditional flow matching. Unlike previous approaches that assume spatially aligned inputs or only handle small-baseline rectified pairs, we address a more challenging setting where severe target noise and large-range two-dimensional cross-modal misalignment coexist. Specifically, we formulate guided denoising as a progressive flow matching process, allowing image restoration and aligned guidance aggregation to mutually reinforce each other. To support efficient few-step inference, we introduce a stage-focused and inference-consistent training strategy that focuses on task-critical denoising stages, and mitigates the training-inference mismatch. We further propose adaptive sparse guidance propagation for efficient and reliable large-range matching, regularized by content and geometry level auxiliary losses. Experiments demonstrate that UGD-FM achieves state-of-the-art denoising performance with only one or two inference steps, and is compatible with modern single-image restoration backbones.

While recent autoregressive video diffusion models achieve remarkable streaming quality, they remain confined to low resolutions (e.g., $480$P), leaving efficient, scalable, real-time high-resolution video generation a fundamental open challenge. To bridge this gap, we present Ultra Flash, a cascaded streaming framework capable of real-time high-resolution video generation. Ultra Flash achieves ${\sim}30$ FPS at 1K resolution and ${\sim}18$ FPS at 2K resolution on a single B200 GPU through three key contributions: (1) an architecture-preserving T2V-to-TV2V super-resolution training paradigm coupled with an AIGC-oriented data degradation pipeline that effectively preserves the generative capability of the base model, enabling enhanced high-resolution detail when cascaded after mainstream low-resolution generative models; (2) a causal streaming latent upsampler paired with a high-resolution decoder, which enhances spatiotemporal coherence while enabling efficient latent spatial scaling and precise high-resolution decoding with negligible computational overhead; and (3) a cascade high-resolution streaming video generation optimization scheme that first performs hybrid-reward-enhanced sparse causalization and single-step distillation of the super-resolution model, then introduces cascaded streaming self-forcing preference optimization with dynamic cache management, jointly enhancing overall coherence, improving quality, and enabling real-time high-resolution streaming video generation. Extensive experiments demonstrate that Ultra Flash reliably produces ultra-high-resolution streaming video while maintaining state-of-the-art visual quality and superior efficiency.

Probabilistic weather forecasting requires not only accurate trajectories, but calibrated distributions over plausible atmospheric futures. Recent data-driven systems have achieved remarkable deterministic skill, and diffusion-based ensemble forecasters have substantially improved sample realism and uncertainty quantification. However, their inference cost scales with forecast horizon, ensemble size, and the number of denoising steps required for each transition, making large operational ensembles expensive. To address this, we present Tyche, a one-step conditional flow model for efficient probabilistic weather forecasting. Tyche models the conditional forecast distribution with a destination-aware average-velocity flow that maps Gaussian noise directly to future weather states in a single function evaluation (1-NFE). To make this one-step transport learnable in high-dimensional geophysical fields, we derive a JVP-regularized rectification objective that enforces temporal self-consistency across source and destination flow timesteps without explicitly forming Jacobians. The transport field is parameterized by an isotropic Swin-style transformer that preserves fine-scale spatial structure while remaining scalable on global grids. To improve ensemble reliability under autoregressive forecasting, we further introduce a rollout-based finetuning stage with curriculum CRPS calibration supervision. Experiments on ERA5 at 1.5$^\circ$ and 6-hour resolution show that our Tyche, using merely a single NFE, matches or exceeds the forecast skill and calibration of state-of-the-art multi-step generative baselines and the operational ECMWF IFS ensemble. Codes are available.

Two-sided matching platforms rely on preferences from both sides, yet participants can evaluate only a small fraction of potential partners. In practice, they use low-cost pre-match screening, e.g., interviews, profile views, or trial tasks, to form noisy impressions before committing to applications and offers. We study bandit learning in matching markets with interviews, modeling these interactions as queried *hints* [Bhaskara et al., 2022] that reveal partial preference information to both sides while constraining subsequent applications. Our framework also allows firm-side uncertainty: firms, like agents, learn their preferences and may make early hiring mistakes. To address this, we introduce strategic deferral, a firm-side action that permits temporary vacancy, corrects premature commitments, and enables decentralized learning under coarse anonymous feedback. We design algorithms for centralized and decentralized markets and show that a constant number of interviews per round suffices for horizon-independent regret, improving over the $O(\log T)$ guarantees known without interviews. Our bounds are near-optimal: the centralized guarantee is within a factor $m$ of an information-theoretic lower bound, while decentralized algorithms match it up to polynomial factors in structured markets and remain horizon-independent in general markets.

Latent generative models have emerged as powerful priors for solving inverse problems. These models typically represent a class of natural signals at a single, fixed complexity, governed by the latent dimensionality. This can be limiting: depending on the problem, a latent dimensionality that is too small may result in high representation error, while one that is too large may overfit to noise. We develop tunable latent priors for diffusion models, normalizing flows, and variational autoencoders, leveraging nested dropout. Across tasks including compressed sensing, inpainting, denoising, and phase retrieval, we show empirically that tunable priors consistently achieve lower reconstruction errors than fixed-complexity baselines. In the linear denoising setting, we derive the optimal complexity in closed form, showing how it depends on the noise level and the signal spectrum. This work demonstrates the potential of tunable latent generative priors and motivates both the development of supporting theory and their application across a wide range of inverse problems.

Reliable map-free LiDAR relocalization requires not only accurate and efficient pose estimation but also trustworthy confidence estimation for detecting localization failures. However, estimating reliability to ensure trustworthy LiDAR relocalization remains largely unexplored. Also, existing LiDAR pose regression methods suffer from accuracy and efficiency trade-off. To address this limitation, we propose \textbf{SPAR}, a model-agnostic confidence estimator that reformulates localization reliability as scan-pose alignment. Given a query scan and a predicted pose, SPAR measures their compatibility in a learned embedding space and explicitly optimizes confidence scores to correlate scan with pose. In addition, we introduce \textbf{VeLoc}, an efficient and accurate pose regressor based on flow-matching for efficient pose refinement. Experiments on the Oxford and NCLT datasets demonstrate that the proposed framework achieves a favorable balance between efficiency and accuracy for LiDAR relocalization while enabling reliable localization failure detection across different localization backbones.

Many problems in diverse fields involve data that naturally live on Lie groups. However, existing Riemannian and Lie group generative models still face several limitations. Riemannian flow and consistency models often require position-dependent vector fields, and expensive geometric terms such as Euler-Arnold simulation on general Lie groups or covariant derivatives. Existing Lie group flow models are usually tied to collinear exponential paths, while momentum-based Lie group diffusion relies on compactness assumptions and does not naturally extend to non-compact groups. To address these limitations, we propose Trivialized Generative Models (TGM), a family of generative models that learns endpoint-constrained paths in the fixed Lie algebra and lifts them to the Lie group. This yields Trivialized Flow Matching (TFM), Trivialized Consistency Models (TCM), and momentum-based extensions with endpoint correction, enabling flexible path design, simpler few-step objectives, and generation on non-compact Lie groups. We evaluate TGM on compact SO(3) benchmarks, non-compact $\mathrm{Sp}(4,\mathbb{R})$ datasets motivated by continuous quantum systems and paraxial optics, and real-world protein backbone generation. Across these settings, TGM improves sample quality and efficiency over existing Riemannian and Lie group baselines.

Recent video diffusion models achieve strong visual quality and temporal coherence, but still lack coordinated control over scene layout, customized subject identity, and camera/subject motion. We study controllable video generation from three prompts: a first-frame scene image, 3D-aware multi-view subject references, and a motion-driving signal. This setting is challenging because background regions are often trackable under camera motion, while foreground subjects can rotate, self-occlude, and reveal new regions that require identity-consistent appearance. We introduce Tri-Prompting, a video diffusion framework that combines multi-view subject conditioning with dual-conditioned motion control. Tri-Prompting uses XYZ tracking points for visible background motion and low-resolution RGB proxies for foreground subject pose, allowing the model to recover fine appearance from multi-view references while retaining flexibility for plausible subject-scene interactions. An inference-time ControlNet scale schedule further balances motion controllability and visual realism. We evaluate Tri-Prompting against specialized baselines: DaS for motion reconstruction and Phantom for subject-driven generation. Tri-Prompting achieves competitive or improved reconstruction quality, stronger multi-view identity preservation, and better 3D consistency, while enabling 3D-aware subject insertion and in-scene manipulation.

Tree-sliced distances have emerged as scalable alternatives to Sliced Wasserstein distances by projecting probability measures onto tree metric spaces and exploiting closed-form transport on trees. Existing tree-sliced constructions, however, are largely built around first-order Wasserstein geometry. While this choice yields efficient computation, it fixes the discrepancy to an $L^1$-type aggregation of subtree mass imbalances and provides limited control over the geometry used in downstream optimization. We propose Tree-Sliced Orlicz IPM (TS-Orlicz), a tree-sliced framework that replaces the tree-level $W_1$ discrepancy with an Orlicz integral probability metric. Building on the tractable formulation of Orlicz IPMs on trees, TS-Orlicz aggregates Orlicz-induced discrepancies over random tree systems while preserving the scalability of tree-sliced computation. By varying the underlying Orlicz function, the framework recovers $L^p$-type behavior and also supports non-polynomial, tail-sensitive geometries that emphasize large subtree discrepancies. We show that TS-Orlicz preserves key properties of tree-sliced constructions and admits efficient computation via closed-form special cases and one-dimensional scalar optimization. We further extend the framework to probability measures on hyperspheres. Experiments across Euclidean and spherical settings, including gradient flows, self-supervised learning, and diffusion-based generative modeling, show that Orlicz-induced tree-sliced discrepancies achieve competitive or improved performance over sliced and tree-sliced baselines while maintaining low computational cost.

Effective discrete representation learning with VQ-VAE serves as a fundamental component of modern autoregressive generative models. However, it is widely known to suffer from codebook collapse and inefficient code utilization. To address this issue, we propose a simple yet effective alternative to standard vector quantization by formulating neural quantization as a joint modeling and clustering process. Specifically, we learn the latent space using a lightweight flow-matching model alongside training, corresponding to a probability flow ODE that transports an often fixed well-behaved source distribution while preserving probability mass. Based on this, we define the codebook $\mathcal{C}_1$ by transporting even-quantile partitioned prior codes $\mathcal{C}_0$ from the source space through the learned flow, and obtaining the Voronoi centroids under the induced distribution, leading a structured discretization. We evaluated on image tasks, and two tasks require high-precision tokenization, demonstrating scalability and improved expressivity of the codebook with full utilization.

Recent image editing systems have achieved impressive semantic understanding, visual fidelity, and instruction-following ability, while video editing remains substantially more difficult and costly. In this paper, we present a simple alternative to end-to-end video editing: instead of training a monolithic video editor, we transform a strong image editor into a video editor through anchor-based generation. Our key insight is that video editing can be decomposed into two subproblems: editing a sparse set of keyframes and propagating those edits across time. Based on this observation, we propose Anchor-based Video Editing (AVE), a two-stage framework in which a powerful image editor first performs composed editing on selected keyframes, and a motion-guided image-to-video diffusion model then generates the final video by treating the edited keyframes as fixed anchors. This design directly inherits the strengths of modern image editors while avoiding expensive end-to-end video editing training. Experiments on IVEBench and VIE-Bench show that AVE achieves strong performance in instruction following, temporal consistency, and content fidelity. Further ablations reveal that final video editing quality is strongly correlated with the quality of the image editor, suggesting that future progress in video editing may come from stronger image editing foundations and lightweight transfer to video.

Diffusion policies have emerged as a powerful policy class for offline reinforcement learning (RL) due to their ability to model complex, multimodal action distributions through iterative denoising. However, in offline RL, expressivity alone is not enough. Generated actions are supposed to remain within value-reliable regions of the action space to avoid extrapolation error. We find that standard diffusion policies often produce unstable denoising trajectories whose intermediate and final action candidates drift toward weakly supported regions, making critic-guided action selection unreliable and degrading teacher quality for one-step distillation. This issue is particularly pronounced on mixed-quality offline datasets, where suboptimal behaviors amplify trajectory instability. To address this problem, we propose Trajectory Consistency Rectification (TCR), a framework designed to promote denoising stability during both training and inference. At inference time, TCR aggregates locally consistent, high-value action candidates across denoising steps to suppress critic-unreliable outliers. During training, it adaptively reweights denoising steps to improve trajectory-level coherence. We show that this stabilization provides a more reliable teacher for one-step flow-matching distillation. Extensive experiments on D4RL and OGBench show that TCR consistently improves over diffusion-policy baselines, with significant gains in sparse-reward domains like AntMaze and Adroit. Moreover, one-step policies distilled from TCR achieve competitive performance among efficient single-step methods. These results suggest that enhancing the temporal consistency of denoising trajectories is a key ingredient for robust offline diffusion reinforcement learning.

Discrete flow matching generates text by iteratively transforming noise tokens into coherent language, but may require hundreds of forward passes. Distillation uses the multi-step trajectory to train a student to reproduce the process in a few steps. When the student underperforms, the usual explanation is insufficient capacity. We argue the opposite: the trajectory is the bottleneck, not the student. Each training trajectory is built through a chain of blind stochastic jumps with no evaluation of sequence quality; a single bad decision at an early midpoint propagates through subsequent steps, yet the student must imitate the result. Trajectory-Shaped Discrete Flow Matching (TS-DFM) replaces these blind jumps with guided navigation: a lightweight energy compass evaluates candidate continuations at each midpoint, selecting the most coherent. All shaping is training-only; inference cost is unchanged. On 170M-parameter language modeling, the shaped student at 8 steps achieves 32% lower perplexity than the 1 024-step teacher while being 128×faster, with gains consistent across source distributions and three evaluators of increasing scale. TS-DFM achieves the best perplexity of any discrete-generation baseline we compare against, including methods trained on 6×more data or using 5×larger models.

Diffusion models generate samples by denoising along the score of a perturbed target distribution. In practice, one trains a neural diffusion model, which is computationally expensive. Recent work suggests that score matching implicitly smooths the empirical score, and that this smoothing bias promotes generalization by capturing low-dimensional data geometry. We propose moment-matched score-smoothed overdamped Langevin dynamics (MM-SOLD), a training-free interacting particle sampler that enforces the target moments throughout the sampling trajectory. We prove that, in the large-particle limit, the empirical particle density converges to a deterministic limit whose one-particle stationary marginal is a Gibbs--Boltzmann density obtained by exponentially tilting a naive score-smoothed diffusion target. The mean and covariance of this distribution agree with the empirical moments of the training data. Experiments on 2D distributions and latent-space image generation show that MM-SOLD enables fast, robust, training-free sampling on CPUs, with sample fidelity and diversity competitive with neural diffusion baselines.