Diffusion
-
Akule: Fast and Scalable Multi-Robot Motion Planning via Sparse Interaction Diffusion
2026-10-09 19:10:23Manav Vora, Gokul Puthumanaillam, Aditya Penumarti, Jane Shin, Melkior OrnikAbstract
Diffusion models provide expressive trajectory priors for multi-robot motion planning (MRMP), but scaling them to large teams requires coordinating individual motions without expensive joint computation. We introduce Akule, a sparse interaction diffusion framework that augments a frozen single-robot prior with three learned modules: a reusable pairwise correction model $R_{\phi}$, a contextual selector $U_{\psi}$ that determines which corrections to evaluate, and a composer $G_{\omega}$ that combines their effects through context-dependent signed weights. This decomposition separates which interactions merit computation from how their effects should be combined into coordinated denoising updates. We train $R_{\phi}$ using matched pairwise trajectory optimizations, then learn $G_{\omega}$ and $U_{\psi}$ from physically optimized joint trajectories without requiring a prescribed ground-truth interaction graph. The resulting joint proposals are passed to a downstream search-based planner for validation and repair. Across diverse MRMP benchmarks, Akule achieves competitive planning success, complete-planner speedups of up to $25\times$, zero-shot population transfer, and successful planning with up to $256$ robots. Ablations further demonstrate that learned selection and composition reduce pre-repair conflicts and downstream repair effort relative to heuristic selection and uniform summation.
Abstract (translated)
URL
https://arxiv.org/abs/2610.14248
PDF
-
InstantFusion: A Shared Latent Interface for Heterogeneous Image Generators
2026-10-09 17:59:04Xuanlang Dai, Jiazi Bu, Yujie Zhou, Yuhong Liu, Beichen Zhang, Penghui Yang, Ziyu Liu, Yuhang ZangAbstract
Diffusion/flow models have emerged as the de facto paradigms for image generation, giving rise to a diverse family of modern generators with increasingly heterogeneous architectures and representations. Despite their remarkable generation capabilities, the cross-model alignment of latent representations and the consistency of their underlying denoising dynamics remain underexplored. In this work, we uncover an intriguing cross-model compatibility: independently trained generators exhibit similar latent spatial organization, while directly handing off intermediate states between unaligned models can still preserve coarse visual content. Motivated by these observations, we propose InstantFusion, a lightweight shared latent interface that connects the denoising trajectories of heterogeneous image generators. Specifically, to enable flexible cross-model state translation, model-specific Latent AutoEncoders (LAEs) are introduced to project intermediate states from different generators into a shared latent space and decode them into the latent space of a target generator. This shared space is constrained by two complementary objectives: aligning intermediate states derived from the same image at matched noise levels and synchronizing their corresponding local denoising velocities. Once trained, InstantFusion allows denoising to seamlessly transition across frozen generators without requiring a dedicated mapping for every model pair. The versatility of InstantFusion is demonstrated across acceleration, multi-preference composition, cross-architecture on-policy distillation, and cross-model composition, achieving competitive performance across these diverse settings.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13944
PDF
-
Images Talk Back: Guiding MMDiTs by Suppressing Image-to-Text Feedback
2026-10-09 17:58:15Yucheng Yang, Xiaoxiao Sun, Shusheng Yang, Jiakang Chen, Serena Yeung-LevyAbstract
Multimodal diffusion transformers (MMDiTs) allow the evolving image to update internal text representations through joint attention. These representations then guide subsequent image updates, raising a question: does this feedback help correct deviations from the prompt, or preserve them? We find that image-to-text (I2T) feedback can preserve the objects and attributes already represented in the evolving image, including errors. Transplanting I2T messages can rescue failures or disrupt successes. At matched intervention strengths, suppressing the influence of I2T feedback improves generation, whereas opposite controls degrade it. These findings motivate Semantic Feedback Guidance (SFG), which suppresses the influence of I2T feedback by guiding away from an auxiliary branch that strengthens this feedback and weakens text-to-image conditioning. The in-place variant, iSFG, directly suppresses I2T feedback and strengthens text-to-image conditioning, requiring no extra model evaluations. Temporal interventions motivate running the SFG branch during the first 20% of image sampling steps, requiring about $1.2\times$ the base NFE without classifier-free guidance (CFG). Experiments across three image-generation backbones show gains in alignment and preference-model scores. On FLUX, adding SFG to CFG raises GenEval from $0.6661$ to $0.7479$; on SD3.5M, iSFG adds $+1.33$ HPSv2.1 over CFG at zero extra NFE. The control also transfers to text-to-video and image-to-video generation, improving VBench scores. Code is available at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13941
PDF
-
Chronos Policy: Learning Robot Policies on Physical Time
2026-10-09 17:55:20Changhao Wang, Jaden Clark, Shuran SongAbstract
Robot policies often convert sensing and control into fixed-rate observation-action sequences, leaving physical time implicit in sequence position. This assumption breaks when observation rates or delays vary, or when inference latency changes. We introduce Chronos Policy, which makes physical time explicit by conditioning on observation capture times and desired action execution times. We study two ways to incorporate time into a diffusion transformer: adding timestamp embeddings to token features, or encoding physical time in positional encoding through time-aware rotary positional encodings (time-aware RoPE). We find that time-aware RoPE is more robust, allowing attention to depend directly on physical time while keeping token content unchanged. During training, we temporally randomize which observations are sampled and when the action chunk begins, exposing the policy to a wider range of sensing and execution schedules that may occur during deployments. This also enables learning from asynchronous, multi-rate data without forcing all measurements onto a fixed temporal grid. Across simulation and real-world experiments, Chronos Policy improves performance on dynamic tasks, where observation-to-action timing is critical, and on multisensory tasks, where the relative timing between observations matters. On standard quasi-static tasks, it preserves nominal performance while improving robustness to changes in sensing and inference timing.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13929
PDF
-
Next-Concept Modeling: Fast, Parallel Coarse-to-Fine Text Generation
2026-10-09 17:49:45Hermann Kumbong, Ishan S. Khare, Iddah Mlauzi, Andrew Woen, Charles Ding, E. Kelly Buchanan, Simran Arora, Azalia Mirhoseini, Christopher R\'eAbstract
Diffusion language models (DLMs) offer parallel sampling for faster generation than autoregressive models, an increasingly important advantage as inference becomes a dominant computational workload. However, DLMs degrade quickly with fewer sampling steps, process the full sequence at every step, and require many passes over the dataset to reach good quality. We introduce Next-Concept Modeling (NCM), a two-stage, coarse-to-fine framework: a multiscale residual VQ-VAE encodes text into a hierarchy of discrete codes, and a transformer generates one level at a time, predicting all codes within a level in parallel. At matched parameter scales, NCM achieves better generative perplexity than the strongest DLM baseline on LM1B while running $21\times$ faster. On OpenWebText, NCM is competitive with the strongest DLM while running $10\times$ faster; on XSum, it matches the strongest DLM's ROUGE-L while being $31\times$ faster. On the reasoning task of Sudoku, NCM outperforms autoregressive and diffusion baselines across all difficulty levels, using fewer function evaluations. On the reasoning task of Sudoku, NCM outperforms autoregressive and diffusion baselines across all difficulty levels, using fewer function evaluations.
Abstract (translated)
URL
https://arxiv.org/abs/2610.14162
PDF
-
HorizonNav: Geometry-Guided Video Diffusion for Anticipatory 3D Robot Navigation
2026-10-09 16:57:08Hajira Amjad, Valerii Serpiva, Viacheslav Liubimov, Didar Seyidov, Dzmitry TsetserukouAbstract
We present HorizonNav, a video diffusion-based framework for anticipatory robot navigation that transforms a single RGB image and a text prompt into a geometrically aligned representation of plausible surrounding 3D space. Given the initial observation and a textual description of the reconstruction objective, a vision-language model, guided by designed instructions, interprets the scene and constructs structured prompts for a video diffusion model. The resulting video sequence provides plausible future observations of the environment, which are subsequently processed using monocular ORB-SLAM3 for camera pose estimation and Video Depth Anything for dense depth prediction. Our central contribution is a geometry-guided depth alignment stage that combines the estimated camera poses with triangulated visual landmarks to constrain dense depth predictions across multiple views. This alignment improves the geometric consistency of the reconstructed 3D environment, addressing inconsistencies that temporal coherence of the generated video alone cannot resolve. The experimental evaluation of navigation in imagined 3D environment shows that across a 19-scene horizon study, planning-feasible scenes increase from 13/19 at P20 to 15/19 at P81, while successful RRT trials increase from 62/90 to 75/90. In a separate corridor study, 6 of 25 evaluable generated-camera trajectories violate the reconstructed collision constraints; RRT planner recovers feasible alternatives in 5 of those 6 cases. A larger archive contains 58 usable reconstructions from 67 attempted cases, of which 36 support a downstream plan and 22 do not, exposing reconstruction and free-space failure modes rather than hiding them without lidar. Across 15 corridor reconstructions, the mean reconstruction error is approximately 0.30 m at matched structural landmarks.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13849
PDF
-
Spanning Diffusion and Flow-based Posterior Sampling via Exact Gaussian Conditioning
2026-10-09 16:24:23Teo Deveney, Lea Bogensperger, Hok Shing WongAbstract
Flow-matching and diffusion models are increasingly used as priors for linear inverse problems, however, most methods are bespoke to one class of trajectory prior. Plug-and-play methods have been proposed, however these implicitly decorrelate the guided trajectory, resulting in broad priors that can struggle when data is sparse. This work proposes $\beta$-Flow: a general plug-and-play method indexed by a correlation parameter $\beta$ that generalises over flow, diffusion, and decorrelated priors, and a confidence parameter $\kappa$ that controls how strongly the path distribution is enforced. Each step requires only one forward evaluation of a denoising network, followed by a Gaussian conditioning step. This step is an exact Kalman update that is not dependent on network gradients. We show that many existing guided samplers can be interpreted as instances or approximations of our method, each implicitly using a different trajectory prior and operator in place of the Kalman gain. From this viewpoint, we provide a qualitative comparison of methods in terms of their implicit biases and assumptions. We demonstrate consistently state-of-the-art level performance across 7 inverse problems over 4 datasets in terms of PSNR and SSIM. Moreover, we demonstrate the utility of this prior control, by showing that confident high correlations (strict priors) work best when observations are less informative, while low correlation is preferred when observations are more informative, reflecting the complementary roles of the prior and likelihood in Bayesian inference.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13812
PDF
-
Your Discrete Graph Diffusion Model is Secretly Memorizing
Abstract
Discrete graph diffusion models (GDMs) corrupt the graph structure by adding noise independently to nodes and edges under a Markovian noise model. In the unconditional setting without node features, an equivariant denoiser trained on a single graph is able to generate novel graphs and sample-level metrics such as edge overlap and WL kernel similarity find no evidence of memorization. In this work, we challenge this view and show for the first time that graph diffusion models indeed exhibit memorization. Analyzing the reverse diffusion trajectories of GDMs trained on two-block stochastic block models (SBMs), we find a strong train-data bias: the model recovers a corrupted training graph almost perfectly compared to a held-out test graph. We term this conditional memorization, as it surfaces only when the reverse trajectory is conditioned on a corrupted training graph and is not apparent during inference. It coincides with a growing generalization gap, where the training loss improves as the loss on held-out graphs degrades, and it cannot be mitigated by early stopping. The onset of memorization always precedes the emergence of community structure in the generated graphs, thus every checkpoint earlier produces structureless graphs. This behavior holds across all node sizes we test.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13786
PDF
-
When Edge Independence Fails: Joint Graph Diffusion with Latent Sociability Priors
2026-10-09 15:50:48Adarsh Jamadandi, Nicolas Keriven, Aline RoumyAbstract
Discrete graph diffusion models corrupt graph structure through independent edge updates with a Markovian noise model. Combined with the exchangeability requirement on the output distribution, the terminal prior is necessarily an ErdÅs--Rényi (E-R) random graph. This has a fundamental consequence for unconditional graph generation without node features: in the large graph limit, any GNN denoiser applied to an E-R graph approximately outputs an E-R graph, making effective denoising difficult. In practice, generated graphs fail to recover meaningful node heterogeneity and higher-order structure. Previous models introduced specialized node features to compensate, framing this as an expressivity problem, while the role of the prior remained unexamined. We instead address the root cause by introducing the Latent Sociability Prior (LSP) and a join} diffusion process that co-evolves latent node sociability variables and graph structure, removing edge independence while preserving exchangeability. Unlike existing latent graph diffusion methods, the sociability mechanism requires no complex encoder: its sole role is to guide edge updates and preserve node heterogeneity throughout the trajectory. We validate our approach on multiple graph generation datasets and show that the proposed framework more accurately captures graph distributions than baseline methods without node features.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13765
PDF
-
SmolWAM: A Latent World-Action Model for Robotic Manipulation on Edge Devices
2026-10-09 15:34:33Timi Owolabi, Hikmah Olawore, Frances Adelakun, Ayotomiwa OyewumiAbstract
World-action models (WAMs) predict a robot's actions together with the observations those actions will produce. However, the strongest models rely on video diffusion transformers that are far too computationally expensive for resource-constrained edge devices. We present SmolWAM, a latent WAM that we derive from the 2B-parameter Cosmos Policy by keeping its pretrained trunk and rebuilding its inference interface: single-frame visual latents, cached text features that preserve the trunk's attention sinks, 12 of 28 transformer blocks, and 1,008 tokens from which actions and future visual latents are read out in one rectified-flow step. Trained in 3.6 GPU-hours, SmolWAM succeeds in 95.5% of 400 LIBERO episodes at 57 ms and 2.3 TFLOP per 16-action chunk with 2.4 GiB of GPU memory, 11x faster than its teacher with 18.5x fewer FLOPs and 3.2x less memory, and its world head still forecasts scene motion.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13740
PDF
-
Improving Semantic Coverage in AI Test & Evaluation through Generative Augmentation
2026-10-09 14:57:12Connor Hashemi, Bharadwaj Ravichandran, Connor Anderson, Shitij Govil, Brandon RichardWebster, Brian Hu, Keith FieldhouseAbstract
In test and evaluation (T&E) of AI models, evaluators typically have access only to a held-out test dataset and no knowledge of the training data or its semantic composition. If critical input conditions---such as time-of-day, weather, or scene location---are underrepresented in this test dataset, T&E results can be misleadingly due to this semantic coverage gap. We propose a three-stage generative AI pipeline to detect and fill such semantic gaps without requiring ground-truth labels of these semantics. First, vision-language models (VLMs) perform zero-shot binary labeling of semantic factors, and an embedding-space clustering step identifies which semantics most affect downstream model behavior. Second, diffusion-based image-to-image models transform existing covered images into the underrepresented semantic domain, preserving scene content while augmenting high-level appearance. Third, statistical analysis of the augmented image embeddings verifies that the synthetic imagery improves dataset coverage before it is incorporated. Crucially, the synthetic data is used only to enrich the test set and no model retraining is performed for T&E. We validate the full pipeline on both image classification and object detection under day/night semantic gaps, showing that the augmented test datasets better distinguish models with and without exposure to the missing semantic, thereby yielding T&E conclusions that more closely match those of an ideal, fully covered dataset. Code available at this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2610.13697
PDF
-
PhysForge: Material-Controllable 3D Reconstruction with Separable Components
2026-10-09 14:47:56Yuchen Wang, Yanqi Bao, Zheng Gu, Zhengqing Chen, Jieqi Shi, Jing HuoAbstract
Recent advances in physical 3D generation have enabled visually faithful assets with simulatable mechanical behavior. However, existing methods provide limited control over material behavior beyond appearance-based inference and struggle to estimate physical properties from complex interactions between objects and semantic parts. Therefore, we present PhysForge, a Video Diffusion Model (VDM)-based framework for addressing these challenges, comprising two complementary modules for material-controllable physical fitting and componentlevel interaction modeling. The first, PhysFit, leverages material-conditioned VDM dynamics as references for physical-property fitting, introducing windowed fitting to focus on material-revealing interactions and Hallucinated Forcing to disentangle nonphysical motion hallucinated by the VDM from material responses. The second, PhysComp decomposes complex interactions into individual objects and semantic components, enabling their physical parameters to be optimized independently and flexibly composed for joint simulation. Experiments on PhysAssets and Google Scanned Objects demonstrate that, compared with existing methods, PhysForge provides stronger control over material-dependent physical responses while supporting component-level physical modeling for complex interactions. It achieves the best CLIP and physical controllability scores on both common and uncommon materials, with qualitative comparisons, component-level ablations, and a user study further supporting these improvements. Our project page is available at this https URL .
Abstract (translated)
URL
https://arxiv.org/abs/2610.13681
PDF
-
RARA: Robust Policy Fine-Tuning through Retrieval-Anchored Representation Alignment for Robotic Manipulation
2026-10-09 13:31:20Xuhui Kang, Sung-Wook Lee, Yan Zhuang, Tianle Zhong, Yen-Ling KuoAbstract
Pre-trained robot policies, such as Vision-Language-Action (VLA) models, provide a strong foundation for generalist robotic manipulation, but adapting them to new tasks with only a few demonstrations remains challenging. Directly fine-tuning a pre-trained policy often achieves strong in-distribution performance, yet it can distort the representations learned during pre-training and weaken the model's generalization ability under out-of-distribution shifts. We propose Retrieval-Anchored Representation Alignment (RARA), a two-stage adaptation method, to address this issue. First, RARA retrieves representation features from pretraining with similar task intent and motion, and aligns the policy's representations of the target demonstrations with the pre-trained representations of the retrieved demonstrations. Second, RARA fine-tunes the policy on the target demonstrations while limiting how much its representations of the retrieved demonstrations change during fine-tuning. By reducing the representation mismatch before action learning and preserving pre-trained knowledge during adaptation, RARA limits unnecessary representation drift during fine-tuning. Experiments with Diffusion Policy and GR00T on 13 tasks in RoboTwin, RoboCasa, and on a real robot show that RARA improves the average success of fine-tuning by 9.5 points in-distribution and 13.6 points out-of-distribution, and outperforms all compared baselines. Project page: this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2610.13578
PDF
-
MOSAIC: Multi-Scale Object-Centric Slot Adaptation with Image-Guided Prototype Conditioning for Remote Sensing Image Super-Resolution
2026-10-09 13:03:57Liuqian Wang, Hanyu Xie, Rongzhen Zhao, Qi Ming, Chunjie Ma, Wentao Chao, Yufei Gao, Chengming Liu, Lei ShiAbstract
Pretrained diffusion Transformers provide powerful generative priors for image restoration, yet adapting them to structure-sensitive tasks remains challenging. Existing methods mainly rely on dense low-resolution features as conditions, which often entangle salient objects with dominant backgrounds and provide limited abstraction for recurring structures across scales and samples. We propose MOSAIC, an object-centric representation learning framework that learns compact structural conditions for generative model adaptation. MOSAIC organizes degraded observations into structural slots that capture complementary object-level and layout-level evidence, integrates them across resolutions, and retrieves recurring priors from a shared prototype memory to complement incomplete image-specific information. The resulting representations are injected into selected layers of a frozen diffusion Transformer through lightweight residual modulation, preserving its pretrained generative prior. We instantiate MOSAIC for remote sensing image super-resolution, where small objects, repetitive structures, and large-scale variation provide a challenging testbed. Experiments on UCMerced, AID, and RSSCN7 show consistent PSNR improvements across $\times 2$ and $\times 4$ settings and favorable LPIPS results at $\times 4$. Qualitative results further demonstrate improved preservation of small objects, structural boundaries, and repetitive spatial patterns.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13547
PDF
-
PartiGen: Part-Aware Representations for Language-Conditioned Humanoid Motion Generation
2026-10-09 10:46:40Chuxuan Huang, Zhanxiang Cao, Tianyue Zhan, Haoyang Li, Yang Zhang, Liyun Yan, Chenlin Ming, Keqi Zhu, Yizhi Chen, Shengcheng Fu, Zihao Zhou, Xinyi Peng, Kaixuan Zhang, Boyang Xing, Jiawei Bao, Yang Yang, Yue GaoAbstract
Translating language into humanoid motion requires coordinating body parts, determining motion duration, and producing references suitable for robot execution. We introduce PartiGen, a part-aware latent diffusion framework that incorporates anatomical structure into a shared whole-body motion representation. Part-specific attention processes anatomically masked frame features, and its outputs are fused to combine local limb information with whole-body context. Conditioned on text and motion history, the model generates motion primitives autoregressively. Forward-kinematics supervision encourages kinematic consistency, while a learned end-of-sequence predictor selects a frame-level endpoint after generation. On HumanML3D retargeted to the Unitree G1, PartiGen achieves a segment-level FID of 3.165, a 16.8% reduction over TextOp, and improves R@1 by 1.4 percentage points. In the duration study with 16-frame primitives, endpoint prediction reduces mean absolute duration error from 12.471 to 3.080 frames compared with a fixed-length cutoff. We further evaluate robot execution using a pretrained, frozen GRIT tracker: 409 of 500 generated references (81.8%) pass the posture-based stability screen in MuJoCo. Following simulation screening and manual selection, 26 of 27 attempted motions are judged successful on the real G1.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13373
PDF
-
Why DiT Models Underperform as Representation Learners without Long Skip Connections
2026-10-09 10:19:11Benyuan Meng, Yiliang Zhang, Jin-Wen Wu, Qianqian Xu, Qingming Huang, Longtao HuangAbstract
Diffusion feature, i.e., intermediate activations extracted from diffusion model backbones, has emerged as a promising approach for unifying generation and vision understanding (discrimination). Despite the advancement of diffusion backbones into Diffusion Transformers (DiT), it has been found in previous literature that DiT models tend to yield weaker features despite their stronger generation capability. This poses a threat to good generation-understanding unification. In this work, we show via controlled from-scratch training that long skip connections (LSCs) affect feature quality: a DiT with LSCs can produce better features. Furthermore, through detailed diagnosis, we reveal the mechanism causing this effect. Specifically, the input to a diffusion backbone contains both the latent image and the diffusion noise, where the diffusion noise, though appearing in the form of noises, actually reflects the information of the diffusion trajectory. To make the final prediction of a noisy variable ($v$ or $\epsilon$), the diffusion backbone needs both signals. However, carrying both signals through the main backbone effectively "pollutes" the image signal with noises, leading to poor feature quality. In comparison, LSCs can shortcut diffusion noises directly to where they are needed, leaving much weaker noise signals in the main backbone, thus improving feature quality. To name the broader impact, this reveals that LSCs are not a mere trick but closely related to the diffusion objective. The code is at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13344
PDF
-
Exploring Attention Sinks in Text-to-Image Transformers from Diffusion and Autoregressive Models
2026-10-09 10:10:51Binglei Li, Mengping Yang, Zhiyu Tan, Wenxu Qian, Junping Zhang, Hao LiAbstract
Attention sinks are tokens or positions that consistently receive disproportionate attention mass, often beyond what their semantic content would suggest. Despite extensive study in large language models, their behavior and functional relevance in text-to-image (T2I) Transformers remain largely unexplored. To fill this gap and better understand how text-to-image Transformers allocate and route attention, we explore this phenomenon of text-to-image transformers across bidirectional diffusion models and causal autoregressive models. Specifically, we explore whether attention sinks exist, where they emerge, and how they functionally influence generation. Through systematic attention analysis and targeted interventions, we identify that attention sinks emerge in both bidirectional diffusion and autoregressive models but exhibit distinct patterns. The autoregressive model concentrates attention on the first and global-scale tokens, whereas the bidirectional models develop sinks near modality boundaries, within text regions, and at spatial corners. Moreover, these patterns vary across generation stages and are localized to a sparse subset of layers and heads. To assess their functional relevance, we perturb sink-related heads and token positions. It turns out that perturbing sink-ranked heads produces larger average changes than non-sink comparisons, while some sink-position perturbations have limited effects on bidirectional model scores. We further introduce register tokens to redistribute attention sinks and find that several configurations preserve or modestly improve generation performance. Overall, attention sinks are architecture-, inference-stage-, and component-dependent phenomena, providing a basis for mechanistic analysis and principled attention design in text-to-image transformers.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13337
PDF
-
LETHE: Defending against Unauthorized Image-to-Video Generation via Value Diversity Collapse
2026-10-09 10:08:39Boheng Li, Yueming Xing, Yiming Li, Jie Zhang, Tianwei ZhangAbstract
Image-to-video (I2V) diffusion models can animate a single reference image into high-quality, subject-consistent videos, posing threats of misuse as anyone can turn personal photos into fabricated videos without consent. A promising solution is to proactively add an adversarial perturbation to the image before release, such that videos generated from it would be unfaithful or visually corrupted. However, due to an insufficient understanding of what governs how subsequent frames read the reference image in modern I2V models, existing protections struggle to offer strong protection under common perturbation budgets. In this paper, we approach the problem by exploring this underlying reading mechanism. Through a convex geometric analysis, we theoretically identify the spread of the reference value vectors as a key factor governing this reading process, and empirically find that even a mild contraction of this spread can severely degrade the generated video, making it an efficient lever for protection. Guided by this insight, we propose LETHE, a novel protection method that directly disrupts this reading process. LETHE pairs two complementary objectives, where a value collapse objective drives the reference value vectors toward uniformity, while an attention attraction objective draws subsequent frames toward these collapsed tokens to amplify corruption. We further construct two human-centric I2V protection benchmarks and conduct extensive experiments with multiple modern I2V models, which demonstrate the state-of-the-art protection performance and versatility of LETHE.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13333
PDF
-
Masked Diffusion Language Agents for Tool-Integrated Chemical Reasoning
2026-10-09 10:07:37Mengdi Liu, Chenghao Jia, Hong Chang, Shiguang ShanAbstract
Designing tool-integrated reasoning agent systems for complex chemical tasks remains a fundamental challenge. Although recent chemical agents have demonstrated promising performance through the integration of external tools, they are still built on the autoregressive paradigm, which is inherently constrained by causal attention and left-to-right generation. As a result, they often struggle with bidirectional molecular understanding, global coordinated tool planning and multi-source evidence integration. To address this challenge, we introduce ChemDiffAgent, the first masked diffusion language agent for tool-integrated chemical reasoning. It reformulates chemical reasoning as iterative denoising over interleaved reasoning and tool-use trajectories, enabling tool-use decisions to be jointly determined under full-trajectory context and thereby better aligning the agent's decision boundary with its knowledge boundary. We further analyze how this paradigm benefits agentic reasoning through harder infilling subproblems during training and flexible arbitrary-order decoding during inference. Then we develop a two-stage post-training framework, comprising agentic supervised fine-tuning and variance-reduced preference optimization, to enhance tool-use and reasoning capabilities. We also construct a comprehensive benchmark covering six chemistry tasks across both single-turn and multi-turn tool-use scenarios. Under matched budgets and evaluation, ChemDiffAgent consistently outperforms autoregressive agents, achieving an overall improvement of approximately 20%. Notably, our 8B model achieves performance comparable to, in some cases surpassing, significantly larger frontier models, such as GPT 5.5 and Claude Opus 4.7. Code is available at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13330
PDF
-
CoVA: Composing VAE Representations for Generative Learning
2026-10-09 10:03:23Zhaowei Chen, Yikai Ge, Linze Li, Yao Tang, Xiang LiAbstract
Representation alignment with pretrained visual encoders improves diffusion transformer training but requires auxiliary encoders and extra training overhead. Recent approaches instead exploit representations from the generative model's own variational autoencoder (VAE). However, reconstruction-oriented VAE features are not optimized for the semantic discrimination and spatial structure needed for generative learning. We introduce \textbf{CoVA} (\textbf{Co}mposed \textbf{VA}E Representations), a lightweight framework that composes VAE features online into stage-wise guidance for flow-matching training. A small latent composer is jointly trained with class supervision, masked latent reconstruction, and horizontal-flip consistency, yielding semantically discriminative, spatially structured dense features and sample-dependent class posteriors. CoVA aligns intermediate SiT features with the composed representations early in training, then removes the increasingly detrimental alignment and reuses detached composer posteriors for sample-dependent soft label conditioning. We further uncover a hidden confound in soft label conditioning: averaging class embeddings attenuates their magnitude, coupling semantic interpolation with conditioning strength. This attenuation alone can substantially improve FID while degrading Recall, revealing a non-semantic source of the apparent gain. CoVA corrects this confound by restoring the hard-label magnitude, isolating posterior-dependent directional semantics. CoVA adds only 8.5M training-only auxiliary parameters (1.3\% of SiT-XL/2) and introduces no inference-time overhead. On ImageNet 256$\times$256, CoVA reduces SiT-XL/2 FID from 17.2 to 7.4 (\textbf{57\%}) in 400K steps. With classifier-free guidance, it achieves an FID of \textbf{2.03} in only 400 epochs, versus 2.06 after 1,400 epochs for vanilla SiT-XL/2.
Abstract (translated)
URL
https://arxiv.org/abs/2610.13322
PDF
