Diffusion Model Guidance

Model-based reinforcement learning (MBRL) with autoregressive world models suffers from compounding errors, whereas diffusion world models mitigate this by generating trajectory segments jointly. However, existing diffusion guides are either policy-only, discarding value information, or reward-based, which becomes myopic when the diffusion horizon is short. We introduce Advantage-Guided Diffusion for MBRL (AGD-MBRL), which steers the reverse diffusion process using the agent's advantage estimates so that sampling concentrates on trajectories expected to yield higher long-term return beyond the generated window. We develop two guides: (i) Sigmoid Advantage Guidance (SAG) and (ii) Exponential Advantage Guidance (EAG). We prove that a diffusion model guided through SAG or EAG allows us to perform reweighted sampling of trajectories with weights increasing in state-action advantage-implying policy improvement under standard assumptions. Additionally, we show that the trajectories generated from AGD-MBRL follow an improved policy (that is, with higher value) compared to an unguided diffusion model. AGD integrates seamlessly with PolyGRAD-style architectures by guiding the state components while leaving action generation policy-conditioned, and requires no change to the diffusion training objective. On MuJoCo control tasks (HalfCheetah, Hopper, Walker2D and Reacher), AGD-MBRL improves sample efficiency and final return over PolyGRAD, an online Diffuser-style reward guide, and model-free baselines (PPO/TRPO), in some cases by a margin of 2x. These results show that advantage-aware guidance is a simple, effective remedy for short-horizon myopia in diffusion-model MBRL.

Classifier-free Guidance (CFG) lets practitioners trade-off fidelity against diversity in Diffusion Models (DMs). The practicality of CFG is however hindered by DMs sampling cost. On the other hand, Consistency Models (CMs) generate images in one or a few steps, but existing guidance methods require knowledge distillation from a separate DM teacher, limiting CFG to Consistency Distillation (CD) methods. We propose Joint Flow Distribution Learning (JFDL), a lightweight alignment method enabling guidance in a pre-trained CM. With a pre-trained CM as an ordinary differential equation (ODE) solver, we verify with normality tests that the variance-exploding noise implied by the velocity fields from unconditional and conditional distributions is Gaussian. In practice, JFDL equips CMs with the familiar adjustable guidance knob, yielding guided images with similar characteristics to CFG. Applied to an original Consistency Trained (CT) CM that could only do conditional sampling, JFDL unlocks guided generation and reduces FID on both CIFAR-10 and ImageNet 64x64 datasets. This is the first time that CMs are able to receive effective guidance post-hoc without a DM teacher, thus, bridging a key gap in current methods for CMs.

Text-to-video diffusion models have enabled open-ended video synthesis, but often struggle with generating the correct number of objects specified in a prompt. We introduce NUMINA , a training-free identify-then-guide framework for improved numerical alignment. NUMINA identifies prompt-layout inconsistencies by selecting discriminative self- and cross-attention heads to derive a countable latent layout. It then refines this layout conservatively and modulates cross-attention to guide regeneration. On the introduced CountBench, NUMINA improves counting accuracy by up to 7.4% on Wan2.1-1.3B, and by 4.9% and 5.5% on 5B and 14B models, respectively. Furthermore, CLIP alignment is improved while maintaining temporal consistency. These results demonstrate that structural guidance complements seed search and prompt enhancement, offering a practical path toward count-accurate text-to-video diffusion. The code is available at https://github.com/H-EmbodVis/NUMINA.

We introduce RewardFlow, an inversion-free framework that steers pretrained diffusion and flow-matching models at inference time through multi-reward Langevin dynamics. RewardFlow unifies complementary differentiable rewards for semantic alignment, perceptual fidelity, localized grounding, object consistency, and human preference, and further introduces a differentiable VQA-based reward that provides fine-grained semantic supervision through language-vision reasoning. To coordinate these heterogeneous objectives, we design a prompt-aware adaptive policy that extracts semantic primitives from the instruction, infers edit intent, and dynamically modulates reward weights and step sizes throughout sampling. Across several image editing and compositional generation benchmarks, RewardFlow delivers state-of-the-art edit fidelity and compositional alignment.

Classifier-Free Guidance (CFG) is a widely used inference-time technique to boost the image quality of diffusion models. Yet, its reliance on text conditions prevents its use in unconditional generation. We propose a simple method to enable CFG-like guidance for both conditional and unconditional generation. The key idea is to generate a perturbed prediction via simple token swap operations, and use the direction between it and the clean prediction to steer sampling towards higher-fidelity distributions. In practice, we swap pairs of most semantically dissimilar token latents in either spatial or channel dimensions. Unlike existing methods that apply perturbation in a global or less constrained manner, our approach selectively exchanges and recomposes token latents, allowing finer control over perturbation and its influence on generated samples. Experiments on MS-COCO 2014, MS-COCO 2017, and ImageNet datasets demonstrate that the proposed Self-Swap Guidance (SSG), when applied to popular diffusion models, outperforms previous condition-free methods in image fidelity and prompt alignment under different set-ups. Its fine-grained perturbation granularity also improves robustness, reducing side-effects across a wider range of perturbation strengths. Overall, SSG extends CFG to a broader scope of applications including both conditional and unconditional generation, and can be readily inserted into any diffusion model as a plug-in to gain immediate improvements.

With the rapid development of conditional diffusion models, significant progress has been made in text-to-video generation. However, we observe that these models often neglect semantically important tokens during inference, leading to biased or incomplete generations under classifier-free guidance. We attribute this issue to two key factors: distributional bias caused by the long-tailed token frequency in training data, and spatial misalignment in cross-attention where semantically important tokens are overshadowed by less informative ones. To address these issues, we propose Distribution-Aware Rectification and Spatial Ensemble (DARE), a unified framework that improves semantic guidance in diffusion models from the perspectives of distributional debiasing and spatial consistency. First, we introduce Distribution-Rectified Classifier-Free Guidance (DR-CFG), which regularizes the training process by dynamically suppressing dominant tokens with low semantic density, encouraging the model to better capture underrepresented semantic cues and learn a more balanced conditional distribution. This design mitigates the risk of the model distribution overfitting to tokens with low semantic density. Second, we propose Spatial Representation Alignment (SRA), which adaptively reweights cross-attention maps according to token importance and enforces representation consistency, enabling semantically important tokens to exert stronger spatial guidance during generation. This mechanism effectively prevents low semantic-density tokens from dominating the attention allocation, thereby avoiding the dilution of the spatial and distributional guidance provided by high semantic-density tokens. Extensive experiments on multiple benchmark datasets demonstrate that DARE consistently improves generation fidelity and semantic alignment, achieving significant gains over existing approaches.

Sequential Monte Carlo (SMC) samplers for reward-guided diffusion models often suffer from rapid lineage collapse: a few high-reward particles dominate the population within a handful of resampling steps, destroying diversity and degrading sample quality. We propose a variance-decomposition framework for reward-guided diffusion SMC that separates continuation variance $V_t^{\mathrm{cont}}$ from residual variance $V_t^{\mathrm{res}}$, revealing that high offspring-count variance under the commonly used multinomial resampling drives this collapse. This motivates VASR (Variance-Aware Systematic Resampling), which addresses both variance terms via variance-optimal mass allocation $m_t \propto w_t e^{r_t}$ (minimizing $V_t^{\mathrm{cont}}$) and systematic resampling (controlling $V_t^{\mathrm{res}}$). For latent diffusion models where intermediate rewards are noisy due to stochastic continuations, we propose VASR-Max, a deliberately biased high-selection variant for variance-sensitive reward optimization. Both methods are training-free, fully parallelizable, and add only linear overhead. On MNIST and CIFAR-10, VASR achieves as high as $26\%$ better FID than prior SMC methods while remaining 66 times faster than MCTS-based value methods at matched compute. On text-to-image generation, VASR-Max consistently outperforms the strongest SMC baseline across compute budgets and matches MCTS-based methods within 2.5--3% reward at high budgets while being approximately times faster.

Diffusion large language models (dLLMs) are emerging as promising alternatives to autoregressive (AR) LLMs. Recently, this paradigm has been extended to multimodal tasks, leading to the development of diffusion multimodal large language models (dMLLMs). These models are expected to retain the reasoning capabilities of LLMs while enabling faster inference through parallel generation. However, when combined with Chain-of-Thought (CoT) reasoning, dMLLMs exhibit two critical issues. First, we observe that dMLLMs often generate the final answer token at a very early timestep. This trend indicates that the model determines the answer before sufficient reasoning, leading to degraded reasoning performance. Second, during the initial timesteps, dMLLMs show minimal dependency on visual prompts, exhibiting a fundamentally different pattern of visual information utilization compared to AR vision-language models. In summary, these findings indicate that dMLLMs tend to generate premature final answers without sufficiently grounding on visual inputs. To address these limitations, we propose Position and Step Penalty (PSP) and Visual Reasoning Guidance (VRG). PSP penalizes tokens in later positions during early timesteps, delaying premature answer generation and encouraging progressive reasoning across timesteps. VRG, inspired by classifier-free guidance, amplifies visual grounding signals to enhance the model's alignment with visual evidence. Extensive experiments across various dMLLMs demonstrate that our method achieves up to 7.5% higher accuracy while delivering more than 3x speedup compared to reasoning with four times more diffusion steps.

Instruction-guided image editing has seen remarkable progress with models like FLUX.2 and Qwen-Image-Edit, yet they still struggle with complex scenarios with multiple similar instances each requiring individual edits. We observe that state-of-the-art models suffer from severe over-editing and spatial misalignment when faced with multiple identical instances and composite instructions. To this end, we introduce a comprehensive benchmark specifically designed to evaluate fine-grained consistency in multi-instance and multi-instruction settings. To address the failures of existing methods observed in our benchmark, we propose Multi-Instance Regional Alignment via Guided Editing (MIRAGE), a training-free framework that enables precise, localized editing. By leveraging a vision-language model to parse complex instructions into regional subsets, MIRAGE employs a multi-branch parallel denoising strategy. This approach injects latent representations of target regions into the global representation space while maintaining background integrity through a reference trajectory. Extensive evaluations on MIRA-Bench and RefEdit-Bench demonstrate that our framework significantly outperforms existing methods in achieving precise instance-level modifications while preserving background consistency. Our benchmark and code are available at https://github.com/ZiqianLiu666/MIRAGE.

The emergence of virtual reality has necessitated the generation of detailed and customizable 3D hand models for interaction in the virtual world. However, the current methods for 3D hand model generation are both expensive and cumbersome, offering very little customizability to the users. While recent advancements in zero-shot text-to-3D synthesis have enabled the generation of diverse and customizable 3D models using Score Distillation Sampling (SDS), they do not generalize very well to 3D hand model generation, resulting in unnatural hand structures, view-inconsistencies and loss of details. To address these limitations, we introduce HandDreamer, the first method for zero-shot 3D hand model generation from text prompts. Our findings suggest that view-inconsistencies in SDS is primarily caused due to the ambiguity in the probability landscape described by the text prompt, resulting in similar views converging to different modes of the distribution. This is particularly aggravated for hands due to the large variations in articulations and poses. To alleviate this, we propose to use MANO hand model based initialization and a hand skeleton guided diffusion process to provide a strong prior for the hand structure and to ensure view and pose consistency. Further, we propose a novel corrective hand shape guidance loss to ensure that all the views of the 3D hand model converges to view-consistent modes, without leading to geometric distortions. Extensive evaluations demonstrate the superiority of our method over the state-of-the-art methods, paving a new way forward in 3D hand model generation.

Controlling the behavior of text-to-image generative models is critical for safe and practical deployment. Existing safety approaches typically rely on model fine-tuning or curated datasets, which can degrade generation quality or limit scalability. We propose an inference-time steering framework that leverages gradient feedback from frozen pretrained foundation models to guide the generation process without modifying the underlying generator. Our key observation is that vision-language foundation models encode rich semantic representations that can be repurposed as off-the-shelf supervisory signals during generation. By injecting such feedback through clean latent estimates at each sampling step, our method formulates safety steering as an energy-based sampling problem. This design enables modular, training-free safety control that is compatible with both diffusion and flow-matching models and can generalize across diverse visual concepts. Experiments demonstrate state-of-the-art robustness against NSFW red-teaming benchmarks and effective multi-target steering, while preserving high generation quality on benign non-targeted prompts. Our framework provides a principled approach for utilizing foundation models as semantic energy estimators, enabling reliable and scalable safety control for text-to-image generation.

Long-range human movement generation remains a central challenge in computer vision and graphics. Generating coherent transitions across semantically distinct motion domains remains largely unexplored. This capability is particularly important for applications such as dance choreography, where movements must fluidly transition across diverse stylistic and semantic motifs. We propose a simple and effective inference-time optimization framework inspired by diffusion-based stochastic optimal control. Specifically, a control-energy objective that explicitly regularizes the transition trajectories of a pretrained diffusion model. We show that optimizing this objective at inference time yields transitions with fidelity and temporal coherence. This is the first work to provide a general framework for controlled long-range human motion generation with explicit transition modeling.

Modern Text-to-Image (T2I) diffusion models have achieved remarkable semantic alignment, yet they often suffer from a significant lack of variety, converging on a narrow set of visual solutions for any given prompt. This typicality bias presents a challenge for creative applications that require a wide range of generative outcomes. We identify a fundamental trade-off in current approaches to diversity: modifying model inputs requires costly optimization to incorporate feedback from the generative path. In contrast, acting on spatially-committed intermediate latents tends to disrupt the forming visual structure, leading to artifacts. In this work, we propose to apply repulsion in the Contextual Space as a novel framework for achieving rich diversity in Diffusion Transformers. By intervening in the multimodal attention channels, we apply on-the-fly repulsion during the transformer's forward pass, injecting the intervention between blocks where text conditioning is enriched with emergent image structure. This allows for redirecting the guidance trajectory after it is structurally informed but before the composition is fixed. Our results demonstrate that repulsion in the Contextual Space produces significantly richer diversity without sacrificing visual fidelity or semantic adherence. Furthermore, our method is uniquely efficient, imposing a small computational overhead while remaining effective even in modern "Turbo" and distilled models where traditional trajectory-based interventions typically fail.

Removing objects from natural images is challenging due to difficulty of synthesizing semantically coherent content while preserving background integrity. Existing methods often rely on fine-tuning, prompt engineering, or inference-time optimization, yet still suffer from texture inconsistency, rigid artifacts, weak foreground-background disentanglement, and poor scalability for multi-object removal. We propose a novel zero-shot object removal framework, namely PANDORA, that operates directly on pre-trained text-to-image diffusion models, requiring no fine-tuning, prompts, or optimization. We propose Pixel-wise Attention Dissolution to remove object by nullifying the most correlated attention keys for masked pixels, effectively eliminating the object from self-attention flow and allowing background context to dominate reconstruction. We further introduce Localized Attentional Disentanglement Guidance to steer denoising toward latent manifolds favorable to clean object removal. Together, these components enable precise, non-rigid, prompt-free, and scalable multi-object erasure in a single pass. Experiments demonstrate superior visual fidelity and semantic plausibility compared to state-of-the-art methods. The project page is available at https://vdkhoi20.github.io/PANDORA.

Transition states, the first-order saddle points on the potential energy surfaces, govern the kinetics and mechanisms of chemical reactions and conformational changes. Locating them is challenging because transition pathways are topologically complex and can proceed via an ensemble of diverse routes. Existing methods address these challenges by introducing heuristic assumptions about the pathway or reaction coordinates, which limits their applicability when a good initial guess is unavailable or when the guess precludes alternative, potentially relevant pathways. We propose to bypass such heuristic limitations by introducing ASTRA, A Priori Sampling of TRAnsition States with Guided Diffusion, which reframes the transition state search as an inference-time scaling problem for generative models. ASTRA trains a score-based diffusion model on configurations from known metastable states. Then, ASTRA guides inference toward the isodensity surface separating the basins of metastable states via a principled composition of conditional scores. A Score-Aligned Ascent (SAA) process then approximates a reaction coordinate from the difference between conditioned scores and combines it with physical forces to drive convergence onto first-order transition states. Validated on benchmarks ranging from 2D potentials to biomolecular conformational changes and a chemical reaction, ASTRA locates transition states with high precision and discovers multiple reaction pathways, enabling mechanistic studies of complex molecular systems.

Motion planning in dynamic urban environments requires balancing immediate safety with long-term goals. While diffusion models effectively capture multi-modal decision-making, existing approaches treat trajectories as monolithic entities, overlooking heterogeneous temporal dependencies where near-term plans are constrained by instantaneous dynamics and far-term plans by navigational goals. To address this, we propose Temporally Decoupled Diffusion Model (TDDM), which reformulates trajectory generation via a noise-as-mask paradigm. By partitioning trajectories into segments with independent noise levels, we implicitly treat high noise as information voids and weak noise as contextual cues. This compels the model to reconstruct corrupted near-term states by leveraging internal correlations with better-preserved temporal contexts. Architecturally, we introduce a Temporally Decoupled Adaptive Layer Normalization (TD-AdaLN) to inject segment-specific timesteps. During inference, our Asymmetric Temporal Classifier-Free Guidance utilizes weakly noised far-term priors to guide immediate path generation. Evaluations on the nuPlan benchmark show TDDM approaches or exceeds state-of-the-art baselines, particularly excelling in the challenging Test14-hard subset.

Face images captured in real-world low light suffer multiple degradations-low illumination, blur, noise, and low visibility, etc. Existing cascaded solutions often suffer from severe error accumulation, while generic joint models lack explicit facial priors and struggle to resolve clear face structures. In this paper, we propose PASDiff, a Physics-Aware Semantic Diffusion with a training-free manner. To achieve a plausible illumination and color distribution, we leverage inverse intensity weighting and Retinex theory to introduce photometric constraints, thereby reliably recovering visibility and natural chromaticity. To faithfully reconstruct facial details, our Style-Agnostic Structural Injection (SASI) extracts structures from an off-the-shelf facial prior while filtering out its intrinsic photometric biases, seamlessly harmonizing identity features with physical constraints. Furthermore, we construct WildDark-Face, a real-world benchmark of 700 low-light facial images with complex degradations. Extensive experiments demonstrate that PASDiff significantly outperforms existing methods, achieving a superior balance among natural illumination, color recovery, and identity consistency. Code and dataset will be available at https://github.com/IVIPLab/PASDiff.

Diffusion models have demonstrated high-quality performance in conditional text-to-image generation, particularly with structural cues such as edges, layouts, and depth. However, lighting conditions have received limited attention and remain difficult to control within the generative process. Existing methods handle lighting through a two-stage pipeline that relights images after generation, which is inefficient. Moreover, they rely on fine-tuning with large datasets and heavy computation, limiting their adaptability to new models and tasks. To address this, we propose a novel Training-Free Light-Guided Text-to-Image Diffusion Model via Initial Noise Manipulation (LGTM), which manipulates the initial latent noise of the diffusion process to guide image generation with text prompts and user-specified light directions. Through a channel-wise analysis of the latent space, we find that selectively manipulating latent channels enables fine-grained lighting control without fine-tuning or modifying the pre-trained model. Extensive experiments show that our method surpasses prompt-based baselines in lighting consistency, while preserving image quality and text alignment. This approach introduces new possibilities for dynamic, user-guided light control. Furthermore, it integrates seamlessly with models like ControlNet, demonstrating adaptability across diverse scenarios.

Diffusion models generate synthetic images through an iterative refinement process. However, the misalignment between the simulation-free objective and the iterative process often causes accumulated gradient error along the sampling trajectory, which leads to unsatisfactory results and a failure to generalize. Guidance techniques like Classifier Free Guidance (CFG) and AutoGuidance (AG) alleviate this by extrapolating between the main and inferior signal for stronger generalization. Despite empirical success, the effective operational regimes of prevalent guidance methods are still under-explored, leading to ambiguity when selecting the appropriate guidance method given a precondition. In this work, we first conduct synthetic comparisons to isolate and demonstrate the effective regime of guidance methods represented by CFG and AG from the perspective of weak-to-strong principle. Based on this, we propose a hybrid instantiation called SGG under the principle, taking the benefits of both. Furthermore, we demonstrate that the W2S principle along with SGG can be migrated into the training objective, improving the generalization ability of unguided diffusion models. We validate our approach with comprehensive experiments. At inference time, evaluations on SD3 and SD3.5 confirm that SGG outperforms existing training-free guidance variants. Training-time experiments on transformer architectures demonstrate the effective migration and performance gains in both conditional and unconditional settings. Code is available at https://github.com/851695e35/SGG.

In practical AI workflows, complex tasks often involve chaining multiple generative models, such as using a video or 3D generation model after a 2D image generator. However, distributional mismatches between the output of upstream models and the expected input of downstream models frequently degrade overall generation quality. To address this issue, we propose Uni-Classifier (Uni-C), a simple yet effective plug-and-play module that leverages video diffusion priors to guide the denoising process of preceding models, thereby aligning their outputs with downstream requirements. Uni-C can also be applied independently to enhance the output quality of individual generative models. Extensive experiments across video and 3D generation tasks demonstrate that Uni-C consistently improves generation quality in both workflow-based and standalone settings, highlighting its versatility and strong generalization capability.

Predicting future states in uncertain environments, such as wildfire spread, medical diagnosis, or autonomous driving, requires models that can consider multiple plausible outcomes. While diffusion models can effectively learn such multi-modal distributions, naively sampling from these models is computationally inefficient, potentially requiring hundreds of samples to find low-probability modes that may still be operationally relevant. In this work, we address the challenge of sample-efficient ambiguous segmentation by evaluating several training-free sampling methods that encourage diverse predictions. We adapt two techniques, particle guidance and SPELL, originally designed for the generation of diverse natural images, to discrete segmentation tasks, and additionally propose a simple clustering-based technique. We validate these approaches on the LIDC medical dataset, a modified version of the Cityscapes dataset, and MMFire, a new simulation-based wildfire spread dataset introduced in this paper. Compared to naive sampling, these approaches increase the HM IoU* metric by up to 7.5% on MMFire and 16.4% on Cityscapes, demonstrating that training-free methods can be used to efficiently increase the sample diversity of segmentation diffusion models with little cost to image quality and runtime. Code and dataset: https://github.com/SebastianGer/wildfire-spread-scenarios

Generative world models have shown promise for simulating dynamic environments, yet egocentric video remains challenging due to rapid viewpoint changes, frequent hand-object interactions, and goal-directed procedures whose evolution depends on latent human intent. Existing approaches either focus on hand-centric instructional synthesis with limited scene evolution, perform static view translation without modeling action dynamics, or rely on dense supervision, such as camera trajectories, long video prefixes, and synchronized multi-camera capture. In this work, we introduce EgoForge, an egocentric goal-directed world simulator that generates coherent, first-person video rollouts from minimal static inputs: a single egocentric image, a high-level instruction, and an optional auxiliary exocentric view. To improve intent alignment and temporal coherence, we introduce GRAFT, a trajectory-level diffusion refinement method that uses positive and negative rollout distributions, derived from goal, temporal, scene-consistency, and perceptual rewards, to steer the diffusion velocity field toward coherent, goal-complete egocentric simulations. Extensive experiments show EgoForge achieves consistent gains in semantic alignment, geometric stability, and motion fidelity over strong baselines, and performs robustly in real-world smart-glasses experiments.

Text-to-image diffusion models achieve high visual fidelity but surprisingly exhibit systematic failures in numerical control when prompts specify explicit object counts. To address this limitation, we introduce ATHENA, a model-agnostic, test-time adaptive steering framework that improves object count fidelity without modifying model architectures or requiring retraining. ATHENA leverages intermediate representations during sampling to estimate object counts and applies count-aware noise corrections early in the denoising process, steering the generation trajectory before structural errors become difficult to revise. We present three progressively more advanced variants of ATHENA that trade additional computation for improved numerical accuracy, ranging from static prompt-based steering to dynamically adjusted count-aware control. Experiments on established benchmarks and a new visually and semantically complex dataset show that ATHENA consistently improves count fidelity, particularly at higher target counts, while maintaining favorable accuracy-runtime trade-offs across multiple diffusion backbones.

Diffusion-based text-to-image (T2I) models have made remarkable progress in generating photorealistic and semantically rich images. However, when the target concepts lie in low-density regions of the training distribution, these models often produce semantically misaligned or structurally inconsistent results. This limitation arises from the long-tailed nature of text-image datasets, where rare concepts or editing instructions are underrepresented. To address this, we introduce Adaptive Auxiliary Prompt Blending (AAPB) - a unified framework that stabilizes the diffusion process in low-density regions. AAPB leverages auxiliary anchor prompts to provide semantic support in rare concept generation and structural support in image editing, ensuring faithful guidance toward the target prompt. Unlike prior heuristic prompt alternation methods, AAPB derives a closed-form adaptive coefficient that optimally balances the influence between the auxiliary anchor and the target prompt at each diffusion step. Grounded in Tweedie's identity, our formulation provides a principled and training-free framework for adaptive prompt blending, ensuring stable and target-faithful generation. We demonstrate the effectiveness of adaptive interpolation over fixed interpolation through controlled experiments and empirically show consistent improvements on the RareBench and FlowEdit datasets, achieving superior semantic accuracy and structural fidelity compared to prior training-free baselines.

The creation of high-fidelity, customizable 3D indoor scene textures remains a significant challenge. While text-driven methods offer flexibility, they lack the precision for fine-grained, instance-level control, and often produce textures with insufficient quality, artifacts, and baked-in shading. To overcome these limitations, we introduce CustomTex, a novel framework for instance-level, high-fidelity scene texturing driven by reference images. CustomTex takes an untextured 3D scene and a set of reference images specifying the desired appearance for each object instance, and generates a unified, high-resolution texture map. The core of our method is a dual-distillation approach that separates semantic control from pixel-level enhancement. We employ semantic-level distillation, equipped with an instance cross-attention, to ensure semantic plausibility and ``reference-instance'' alignment, and pixel-level distillation to enforce high visual fidelity. Both are unified within a Variational Score Distillation (VSD) optimization framework. Experiments demonstrate that CustomTex achieves precise instance-level consistency with reference images and produces textures with superior sharpness, reduced artifacts, and minimal baked-in shading compared to state-of-the-art methods. Our work establishes a more direct and user-friendly path to high-quality, customizable 3D scene appearance editing.

Bridging the simulation-to-reality (sim2real) gap remains challenging as labelled real-world data is scarce. Existing diffusion-based approaches rely on unstructured prompts or statistical alignment, which do not capture the structured factors that make images look real. We introduce Ontology- Guided Diffusion (OGD), a neuro-symbolic zero-shot sim2real image translation framework that represents realism as structured knowledge. OGD decomposes realism into an ontology of interpretable traits -- such as lighting and material properties -- and encodes their relationships in a knowledge graph. From a synthetic image, OGD infers trait activations and uses a graph neural network to produce a global embedding. In parallel, a symbolic planner uses the ontology traits to compute a consistent sequence of visual edits needed to narrow the realism gap. The graph embedding conditions a pretrained instruction-guided diffusion model via cross-attention, while the planned edits are converted into a structured instruction prompt. Across benchmarks, our graph-based embeddings better distinguish real from synthetic imagery than baselines, and OGD outperforms state-of-the-art diffusion methods in sim2real image translations. Overall, OGD shows that explicitly encoding realism structure enables interpretable, data-efficient, and generalisable zero-shot sim2real transfer.

Retrieval-Augmented Generation (RAG) improves factual grounding by incorporating external knowledge into language model generation. However, when retrieved context is noisy, unreliable, or inconsistent with the model's parametric knowledge, it introduces retrieval-prior conflicts that can degrade generation quality. While this problem has been studied in autoregressive language models, it remains largely unexplored in diffusion-based language models, where the iterative denoising process introduces unique challenges for integrating retrieved context. In this work, we propose Adaptive Retrieval-Augmented Masked Diffusion (ARAM), a training-free adaptive guidance framework for Masked Diffusion Models (MDMs) in RAG settings. ARAM dynamically calibrates the guidance scale during denoising according to the Signal-to-Noise Ratio (SNR) of the distributional shift induced by retrieved context. Intuitively, the model strengthens guidance when the retrieved context provides reliable corrective evidence and suppresses it when the contextual signal is noisy or non-supportive. Extensive experiments on multiple knowledge-intensive QA benchmarks show that ARAM improves overall QA performance over competitive RAG baselines.

Guided diffusion sampling relies on approximating often intractable likelihood scores, which introduces significant noise into the sampling dynamics. We propose using adaptive moment estimation to stabilize these noisy likelihood scores during sampling. Despite its simplicity, our approach achieves state-of-the-art results on image restoration and class-conditional generation tasks, outperforming more complicated methods, which are often computationally more expensive. We provide empirical analysis of our method on both synthetic and real data, demonstrating that mitigating gradient noise through adaptive moments offers an effective way to improve alignment.

In this paper, we propose Iris, a deterministic framework for Monocular Depth Estimation (MDE) that integrates real-world priors into the diffusion model. Conventional feed-forward methods rely on massive training data, yet still miss details. Previous diffusion-based methods leverage rich generative priors yet struggle with synthetic-to-real domain transfer. Iris, in contrast, preserves fine details, generalizes strongly from synthetic to real scenes, and remains efficient with limited training data. To this end, we introduce a two-stage Priors-to-Geometry Deterministic (PGD) schedule: the prior stage uses Spectral-Gated Distillation (SGD) to transfer low-frequency real priors while leaving high-frequency details unconstrained, and the geometry stage applies Spectral-Gated Consistency (SGC) to enforce high-frequency fidelity while refining with synthetic ground truth. The two stages share weights and are executed with a high-to-low timestep schedule. Extensive experimental results confirm that Iris achieves significant improvements in MDE performance with strong in-the-wild generalization.

Do video diffusion models encode signals predictive of physical plausibility? We probe intermediate denoising representations of pretrained Diffusion Transformers (DiTs) and find that physically plausible and implausible videos are partially separable in mid-layer feature space, even at high noise levels. Within-source and perceptual-quality controls suggest that this signal is not fully explained by generator identity or generic visual quality. We distill the signal into a lightweight, backbone-specific physics verifier trained on frozen features and use it in two complementary inference-time mechanisms under a fixed multi-trajectory budget: progressive trajectory selection, which scores trajectories at intermediate checkpoints and prunes weak candidates early, and reward-gradient guidance, which steers surviving trajectories by backpropagating through only the first few DiT blocks. Experiments on PhyGenBench and Physics-IQ across CogVideoX-2B/5B and Wan 2.1-14B show that progressive selection matches verifier-based Best-of-4 on CogVideoX-2B while reducing wall-clock inference time by 37%, whereas reward-gradient guidance substantially improves physical consistency on CogVideoX-5B, all without fine-tuning the video generator.