Denoising
-
Decoupled Early Exits for Task-Dependent Compute Allocation in Flow-Matching VLAs
2026-09-24 11:10:00Riccardo Andrea Izzo, Rimvydas Rubavicius, Gianluca Bardaro, Subramanian Ramamoorthy, Matteo Matteucci, Alessandro SugliaAbstract
Flow-matching Vision-Language-Action (VLA) models have emerged as a potential solution for generalist robot control, designed by combining a pretrained Vision-Language Model (VLM) backbone with an action expert that generates continuous robot actions. While these models exhibit impressive capabilities, due to their very high number of parameters, their computational requirements are often prohibitive for robotics control. To mitigate these inefficiencies, existing methods predominantly skip VLM backbone layers with early exits or reduce denoising steps, while leaving action expert depth untouched. We propose a framework that exposes backbone depth $V$, action expert depth $A$, and denoising steps $D$ as three jointly configurable compute axes in a VLA. Starting from a pretrained VLA, we attach lightweight Exit Transformers (ET) at intermediate depths in both the backbone and the action expert, trained to distil the last layer of the policy into each exit. Furthermore, we introduce a KV Cache synthesis mechanism that manages the missing keys and values of the skipped backbone layers, allowing the action expert to exit deeper than the backbone. Finally, we show that the optimal compute budget is task-dependent, with different tasks benefiting from different axes and depths. Notably, our method does not require training the original policy from scratch, and for each exit, it increases the number of parameters by only $2.1\%$ for SmolVLA and $4.1\%$ for $\pi_{0.5}$. We validate our approach across two flow-matching VLAs (SmolVLA, $\pi_{0.5}$) and two benchmarks (LIBERO, Meta-World), revealing complementary effects: $V$ and $A$ respectively reduce FLOPs and latency, while $D$ improves both. Our joint configurations $(V,A,D)$ reduce latency by $79.2\%$ and computation (FLOPs) by $31.8\%$, while improving mean success rate by $5.6\%$.
Abstract (translated)
URL
https://arxiv.org/abs/2609.29382
PDF
-
A Study of the Limits of Collaborative DCT-Based Image Denoising via Interpretable Neural Networks
2026-09-24 10:07:59Cristian Comellas, Julia Navarro, Antoni BuadesAbstract
Image denoising remains a fundamental problem in image restoration, with applications in photography, biomedical, and scientific imaging. Modern deep neural networks achieve strong performance by learning powerful image priors, but often rely on large black-box models with limited interpretability. In contrast, DCT-based sliding-window and collaborative filtering methods such as BM3D offer clear algorithmic structure, but depend on handcrafted and non-differentiable operations. This work studies how far such structured collaborative filtering principles can be pushed when reformulated as trainable models. We introduce DeepBM3D, a compact fully differentiable architecture that combines non-local patch grouping, DCT-domain filtering, and multi-stage refinement within a BM3D-inspired pipeline. Lightweight convolutional feature extractors guide patch grouping, while filtering is performed through learned Wiener weights in the DCT domain. Experiments show that DeepBM3D improves over classical and hybrid baselines, remains competitive with FFDNet at low and moderate noise levels, and performs particularly well on repetitive textures.
Abstract (translated)
URL
https://arxiv.org/abs/2609.29334
PDF
-
TOLA: Text-aware One-Step Latent Adaptation for Diffusion-based Text Image Super-Resolution
2026-09-24 08:48:10Yike Xu, Yue Shi, Yong Guo, Jiezhang CaoAbstract
Text image super-resolution (TSR) aims to recover visually faithful and readable text under unknown degradations. Existing diffusion-based methods typically rely on multi-step prediction of either the high-resolution image or its text prior, resulting in prohibitive computational cost and inference latency. More critically, an erroneous text prior may be repeatedly injected into the denoising process, causing image and text predictions to reinforce each other and progressively amplify an early recognition error into a sharp yet semantically incorrect character. To address these limitations, we propose TOLA, a Text-aware One-step Latent Adaptation framework without iterative image-text diffusion. TOLA consists of two key modules. First, a confidence-weighted text conditioning module constructs the semantic condition only once and suppresses unreliable OCR predictions before they contaminate image reconstruction. Second, a lightweight latent residual correction module explicitly estimates and corrects the structured residual errors to recover missing or distorted stroke details. Extensive experiments demonstrate our state-of-the-art performance across all evaluation metrics on both CTR-TSR-Test ($\times 4$) and RealCE-200 benchmarks. It is worth noting that our TOLA consistently surpasses existing diffusion-based TSR methods by at least 2.72 dB in PSNR on CTR-TSR-Test.
Abstract (translated)
URL
https://arxiv.org/abs/2609.29240
PDF
-
ComplexSync: High-Fidelity and Real-Time Lip Sync in Complex Scenarios
2026-09-24 08:33:21Jiaran Cai, Xingpei Ma, Shenneng HuangAbstract
Lip synchronization aims to generate visual lip dynamics that align precisely with speech audio. Despite the high generation quality of diffusion models, they often struggle in complex scenarios and suffer from prohibitive inference latency, limiting real-world deployment. We present ComplexSync, a unified diffusion-based framework that enables real-time, high-fidelity lip sync under complex conditions. First, we introduce a dual-stream joint training strategy to mitigate information leakage from reference frames while preserving natural dynamics. Second, we develop a distillation-based acceleration scheme for single-step denoising, achieving a throughput of over 70 FPS. Third, we propose a relational alignment loss that leverages structural priors from Vision Foundation Models (VFMs) to enhance robustness against complex scene factors. Furthermore, we present the first benchmark specifically designed for complex lip synchronization, comprising over 200 challenging video sequences and specialized metrics. Extensive experiments demonstrate that ComplexSync achieves state-of-the-art performance across both standard and complex scenarios while enabling real-time inference.
Abstract (translated)
URL
https://arxiv.org/abs/2609.29225
PDF
-
FB-GDM: Fully-Bayesian Guided Diffusion Models for High-Dimensional Linear Inverse Problems via Unsupervised Variational Inference
2026-09-24 08:24:38Gatien S\'eguy (SATIE), Thomas Rodet (SATIE)Abstract
Diffusion models are powerful priors for linear inverse problems, but the reference guidance methods, Diffusion Posterior Sampling (DPS) and Pseudoinverse-Guided Diffusion Models ($\Pi$GDM), rely on scalar hyperparameters tuned per task, usually against the ground truth. We introduce FB-GDM, a fully-Bayesian guided diffusion method that removes this calibration step. Starting from the Gaussian approximation of $\Pi$GDM, we derive a closed-form conditional score that depends on two precision parameters (inverse variances), one associated with the denoising approximation and one with the observation likelihood, and treat them as latent variables inferred by variational inference at each reverse step. A separable factorization makes each update scale linearly with the number of pixels, so the inference stays tractable at full image resolution, at a cost comparable to one $\Pi$GDM run. FB-GDM requires neither the noise level nor the ground truth: its only inputs are the observation and the forward operator. Experiments on CelebA-HQ inverse problems establish two results. (i) The precision parameters, inferred from the observation alone, allow FB-GDM to outperform $\Pi$GDM at its nominal setting, even when the latter is given the true noise level, by up to 14 dB depending on the operator, and to match the ground-truth-calibrated $\Pi$GDM oracle within 0.1 dB. (ii) FB-GDM is robust when the forward operator, the noise level, or the image distribution changes: it stays close to a per-problem $\Pi$GDM oracle throughout and does not exhibit the hallucinations observed with DPS, whereas DPS substantially degrades at a fixed scale and $\Pi$GDM stays competitive only if it is re-tuned against the ground truth for each new problem. When the prior is applied to images outside its training set, this re-balancing between data and prior keeps FB-GDM faithful where a fixed face-prior guidance can otherwise hallucinate.
Abstract (translated)
URL
https://arxiv.org/abs/2609.29216
PDF
-
ELF-REG: Scaling Continuous Diffusion Language Models to Reasoning Tasks
2026-09-24 06:33:19Zeyu Michael Li, William Xingxu Chen, Bingshuo Qian, Jiayin Liu, Xiang ChengAbstract
Fully continuous diffusion language models (dLMs) denoise continuous representations without intermediate discretization, then decode all response tokens in parallel at the final step. Their performance on challenging reasoning tasks remains less established than that of autoregressive (AR) LLMs and masked dLMs. We scale Embedded Language Flows (ELF) to mathematical reasoning and code generation on GSM8K, MATH-500, HumanEval, and MBPP. We introduce ELF-REG, which improves learning with representation alignment and entanglement (REPA+REG), where a frozen AR teacher supervises intermediate denoiser features and supplies a global representation that is jointly denoised with the response. ELF-REG-L achieves 55.96% pass@1 on GSM8K at 64 network function evaluations (NFE), and 13.39% on MATH-500 and 22.56% on HumanEval at 128 NFE. It outperforms the evaluated comparable-scale dLMs in pass@1 on GSM8K and code, and improves MATH-500 pass@1 from 10.55% for the ELF-L baseline to 13.39% with ELF-REG-L. Without few-step training, the same task-specific checkpoints support strong low-NFE performance through early-stop, which decodes an intermediate clean prediction without completing the denoising trajectory. At 16 NFE, ELF-REG-L reaches 41.21% HumanEval pass@10, outperforming recent continuous dLMs of comparable scale.
Abstract (translated)
URL
https://arxiv.org/abs/2609.29102
PDF
-
DAWN: Noise-Robust Quadruped Parkour via Depth-Denoising World Models
2026-09-24 06:20:19Yohan Choi, Min-Jun Kim, Jin-Sung Kim, Yong-Jae Kim, Youn-Hee HanAbstract
Vision-based legged locomotion methods assume clean depth at training time and rely on hand-tuned post-processing filters at deployment. However, filter parameters are rarely disclosed, hindering reproducibility, and performance degrades substantially when depth noise is left unaddressed. Building noise robustness directly into the learning pipeline would eliminate this dependency. While such robustness has been explored for proprioceptive inputs, analogous approaches for depth perception remain largely absent in legged locomotion. We propose DAWN (Denoising and Alignment in World models for Noise-robustness), a noise-robust perception framework for legged locomotion, which builds noise robustness directly into a world model via two modifications: (1) feeding noisy depth to the encoder while keeping clean depth as the reconstruction target, forcing the model to implicitly denoise its input; and (2) applying contrastive learning to align the latent states of noisy and clean depth. Importantly, DAWN is not tied to a specific noise model, requiring no manual tuning to the noise distribution at deployment. Furthermore, it incurs no additional inference cost over existing world model-based methods. Without any manual filter calibration -- relying solely on the learned noise-robust representation -- DAWN achieves zero-shot quadruped parkour on a Unitree Go1: traversing stairs up to 18 cm, clearing gaps up to 70 cm, and mounting steps up to 45 cm from raw depth observations. Ablation studies show that denoising and contrastive alignment contribute at complementary levels -- reconstruction and representation, respectively -- and yield additive gains when combined. Videos and code are available at: this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2609.29092
PDF
-
The Vulnerability of Neural Audio Watermarks under Speech Enhancement
2026-09-24 05:17:12Xincong Zhong, Shengyao Wang, Lingfeng Yao, Yihang Bao, Jinze Yu, Miao Pan, Jiang LiuAbstract
Neural audio watermarks are increasingly deployed in commercial speech generation systems to make AI-generated speech traceable, yet their robustness has been studied mainly under conventional signal distortions. Since a watermark can be regarded as imperceptible noise added to the speech signal, a natural question is whether speech enhancement (SE), as a denoising model, can remove it. In this paper, we cascade Gaussian noise with SE models as a black-box watermark removal attack, covering both discriminative and generative SE paradigms, against six neural watermarks: AudioSeal, WavMark, SilentCipher, Timbre, Perth, and AlignMark. Experimental results show that the proposed attack significantly outperforms existing neural re-synthesis methods in watermark removal. In particular, we find that generative SE, which reconstructs the harmonic regions of speech while denoising, is highly destructive to watermarks. These findings show that SE poses a serious threat to current audio watermarking methods, and we call for SE-aware robustness evaluation in watermark design.
Abstract (translated)
URL
https://arxiv.org/abs/2609.29040
PDF
-
Beyond Future Prediction: Denoising as Generative Adaptation for Robot Control
2026-09-23 16:17:35Zanyi Wang, Yuheng Lei, Dengyang Jiang, Ping Luo, Mengdi Wang, Zhixuan Liang, Shilong LiuAbstract
Pretrained generative Diffusion Transformers (DiTs) capture rich pixel-level visual and language-conditioned structure through large-scale image and video generation training. A growing line of robot policies builds on this generative prior, but how it should be transferred to control remains unclear, and existing approaches commonly instantiate this transfer through future visual prediction. We ask a more basic question: what a pretrained generative DiT actually contributes to action learning, and how this prior should be adapted for control. We introduce NowWAM, a future-target-free co-training formulation that denoises the current observation and predicts robot actions from the same visual stream, directly coupling the native generative objective to the action-facing representation across the denoising trajectory. Under matched controlled settings, past and future visual targets perform comparably, while restricting training to the clean endpoint substantially reduces robustness, suggesting that a separate future target is not essential for generative adaptation, while the denoising trajectory remains an effective interface for control. On LIBERO-Plus, NowWAM reaches 87.7% with FLUX2-Klein, improving over the future-target co-training baseline by 6.1 points while halving training visual tokens (784 to 392) and reducing step time from 2.85 s to 1.63 s, a 1.8x speedup. With the pure text-to-image Z-Image backbone, NowWAM further reaches 87.8%, showing that strong control adaptation is not tied to video generation or image-editing backbones.
Abstract (translated)
URL
https://arxiv.org/abs/2609.28339
PDF
-
Benchmarking Hyperspectral Foundation Models for Hyperspectral Unmixing
2026-09-23 15:32:44Edgard Dabier, Christophe Kervazo, Pietro Gori, Florence TupinAbstract
Several foundation models dedicated to hyperspectral images have recently been made available. These models are trained on large unlabeled datasets and exhibit strong performance on many hyperspectral imaging tasks, such as classification or denoising. Nonetheless, their performance for hyperspectral unmixing -- the task of separating mixed spectra of overlapping materials in a hyperspectral image -- remain understudied. This might partly be due to the fact that most of them rely on vision transformer backbones, including patchification, leading to a feature resolution problem. While hyperspectral unmixing already arises from the low resolution of hyperspectral images, this patchification step potentially makes the problem even more ill-posed. Therefore, in this work, we aim to answer two questions: 1) \emph{how do foundation models perform in hyperspectral unmixing?}; 2) \emph{how to tackle the feature-level loss of resolution?} To answer the first question, we benchmark foundation models for unmixing, showing that they can reach state-of-the-art performance on four hyperspectral unmixing datasets. To answer the second question, we compare several feature upsampling approaches and empirically show that using a simple one can lead to high performance results. The code is available at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2609.28283
PDF
-
ZoomDiff: A High-Fidelity Diffusion Model for Dual-Camera Smooth Zooming
2026-09-23 13:22:58Jiayi Zhang, Renlong Wu, Yukang Ding, Sibin Deng, Wangmeng ZuoAbstract
Digital zoom transitions between dual cameras often exhibit conspicuous discontinuities in geometric structure and chromatic consistency, degrading the user experience. While recent dual-camera smooth zoom (DCSZ) methods attempt to mitigate this by fine-tuning frame interpolation (FI) models on DCSZ data, they struggle with the large cross-view disparities and complex geometric transformations. Considering that the generative prior of diffusion models is suitable for addressing this problem, we explore their application to DCSZ. However, naively applying existing diffusion-based FI models still yields low-fidelity transitions due to insufficient conditional guidance, high-frequency information loss during VAE encoding, as well as inadequate temporal consistency. To address this, we propose ZoomDiff, a high-fidelity diffusion model that leverages dual-camera inputs in both latent and pixel spaces for photo-realistic transitions. Specifically, we first strengthen dual-image conditional guidance during the multi-step denoising process to improve geometric consistency. Then we inject flow-aligned multi-scale features from the VAE encoder into the VAE decoder to recover high-frequency details, where flow-guided temporal consistency supervision are introduced to produce more smooth transitions. Extensive experiments on both synthetic and real-world datasets demonstrate that ZoomDiff outperforms state-of-the-art methods quantitatively and qualitatively. Project page: this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2609.28083
PDF
-
InGuard: Towards Generalized Inner Guardrail for Safe Text-to-Image Generation
2026-09-23 09:44:09Zeyu Wang, Xiaodan Li, Zhiwen Li, Yuefeng Chen, Hui XueAbstract
Modern text-to-image (T2I) models generate high-quality images from arbitrary user prompts, yet they can just as easily produce not-safe-for-work (NSFW) content. Conventional outer guardrails consist of two components: a prompt classifier that checks for risk before generation, and a post-hoc image classifier that checks the fully generated image. In this design, both classifiers operate outside the generation pipeline and do not use the model's own representations. This separation can limit prompt-screening accuracy, while the image-side check runs only after the full generation cost has been spent. Moreover, a flagged prompt can only be rejected, even when it could be adjusted to produce a safe image. In this work, we propose the Inner Guardrail (InGuard), a safety framework that works inside the pipeline on the model's own representations, leaving base-model parameters untouched. First, a risk classifier grades each prompt as unsafe, risky, or benign based on the text encoder's embeddings, with no external language model. Second, SAGE (Soft-gated Asymmetric Guardrail for Embeddings) modifies the embeddings of risky prompts, aiming to return a safe image instead of a refusal. Third, a latent detector checks the one-step clean latent estimate midway through denoising, reaching nearly image-level performance and halting generation when risk is detected. We also construct the RevGen Safety Benchmark to evaluate T2I safety under realistic conditions: 10,000 prompts built through real-image reverse generation, with a rewriting step that supplies controlled intellectual-property (IP) characters, covering graded porn/gore risks, categorical IP risks, and benign negatives. Across five open-weight T2I models, InGuard reaches 97.9-98.8% safety rate, matching or exceeding the outer guardrail, with 57.5-73.5% less benign disturbance, ~3.7x fewer parameters, and 50-55.6% of denoising steps skipped.
Abstract (translated)
URL
https://arxiv.org/abs/2609.27620
PDF
-
Information Capacity of Generative Video Compression: Quantifying the Rate-Compute Exchange at Identical Quality
2026-09-23 07:56:15Cheng Yuan, Jiawei Shao, Xuelong LiAbstract
Under the AI Flow framework, communication networks distribute intelligence across devices, edge servers, and clouds, and computation at the receiver becomes a resource that can substitute for transmitted bits. Generative video compression (GVC) embodies this exchange by sending compact tokens with ultra-low bitrate and letting a generative decoder synthesize the video, yet how much bandwidth savings a unit of decoder compute actually achieves has never been quantified. To fill this vacancy, we model reconstruction quality as a two-factor power law in data rate and decoder compute, which fits measured DISTS of two GVC decoders with a mean error below 3%, and define the information capacity (IC) as the negative logarithmic slope along an iso-quality contour, namely the fraction of rate saved per fractional increase in compute at identical quality. IC is dimensionless and unit-invariant, thus enabling an architecture-agnostic comparison. It forms a field over the operating plane, locating where additional denoising steps are worth their cost. Across five datasets, the 14B decoder trades more compute for fewer rate about ten times more efficiently than the 1.3B decoder. IC also varies significantly across datasets, indicating imbalanced performance on the rate-compute trade-off in GVC methods.
Abstract (translated)
URL
https://arxiv.org/abs/2609.27493
PDF
-
LOCKR: A Hidden-State Trajectory-Guided Planner for Detecting and Repairing Stable-but-Wrong Lock-In in Diffusion Language Models
2026-09-23 01:35:24Guoshenghui Zhao, Tan Yu, Weijie ZhaoAbstract
Diffusion language models generate text through iterative denoising, exposing intermediate trajectories before final answers are produced. We identify a recurring reasoning failure, stable-but-wrong lock-in, where an answer stabilizes early around an incorrect value while substantial denoising remains. Surface-level decoding signals such as confidence, entropy, margin, and answer stability are insufficient to reliably distinguish correct from erroneous lock-in. We formulate selective reasoning repair as a lightweight test-time planning problem and propose LOCKR, a hidden-state trajectory-guided planner that decides when to allocate additional computation, expands a structured set of targeted repair branches, and selects the most promising continuation using trajectory-aware verification. Across two diffusion language models and three mathematical reasoning benchmarks, hidden-state trajectories consistently outperform surface signals and single hidden snapshots for both wrong-lock-in detection and repair selection. On natural evaluation distributions, LOCKR yields absolute accuracy gains of 2.21--5.37 percentage points across all five evaluated settings, with repair rates ranging from 22% to 41%. These results establish hidden diffusion trajectories as actionable signals for selective test-time reasoning repair.
Abstract (translated)
URL
https://arxiv.org/abs/2609.27220
PDF
-
On Preference Coverage Collapse from Hindsight Relabeling in Multi-Objective Reinforcement Learning
2026-09-22 18:13:01Baptiste Bonin, Caro Strickland, Audrey DurandAbstract
Hindsight relabeling which retroactively replacing a transition's goal with the outcome the agent actually achieved is an effective tool for improving sample-efficiency in Reinforcement Learning (RL). A natural extension to preference-conditioned multi-objective RL (MORL) relabels transitions with the preference direction the agent achieved rather than the one asked for. We show that this extension is frequently harmful: across four preference-conditioned off-policy algorithms spanning two critic backbones and two preference-sampling schemes on the continuous-control MO-Gymnasium suite, it degrades 19 of 36 algorithm-environment settings by as much as four standard deviations, improves only one, and leaves the rest unaffected. The harm is not a symptom of noisy relabels; denoising the target recovers almost nothing, and neither prioritized sampling nor any buffer-structural choice reproduces it. Instead, repeated relabeling collapses the critic's coverage onto whatever narrow region of the preference space the agent happened to visit. We name this failure mode \emph{Preference Coverage Collapse}, and quantify it with abandoned preference mass (APM), a value-aware statistic that tracks the harm ($\rho = -0.73$) where a purely structural coverage count does not. We then introduce \texttt{her\_mix}, a single-parameter convex combination pulling the achieved direction back towards the requested preference. At one fixed value across every algorithm and environment, it returns 16 of the 19 harmed settings to baseline, preserves and even improves the one setting in which relabeling helps, and cuts abandoned preference mass from $69\%$ to $6\%$. Protecting coverage over the preference simplex, not filtering noisy relabels, is what makes hindsight relabeling safe for MORL.
Abstract (translated)
URL
https://arxiv.org/abs/2609.26918
PDF
-
ASTRA-SR: Atmospheric Seeing and Turbulence Restoration for Astronomical Image Super-Resolution
2026-09-22 17:22:02Xining Ge, Ziteng Cui, Shuhong LiuAbstract
Ground-based planetary imaging suffers from atmospheric turbulence, sensor noise, and limited sampling, making restoration a joint denoising, deblurring, and super-resolution problem. We present ASTRA-SR, a blind single-frame restoration framework trained on a physics-grounded synthetic dataset. High-dynamic-range spacecraft RAW observations serve as clean sources, and paired LR inputs are synthesized using measured layer-integrated turbulence strengths, propagated moving phase screens, exposure-averaged spatially varying PSFs, and sensor this http URL-SR first estimates a noise-suppressed but blur-retaining LR image, then restores spatial structure through multiscale processing and reconstructs HR detail with serial spatial-amplitude refinement. It yields a 0.49 dB foreground PSNR gain over the strongest baseline approaches.
Abstract (translated)
URL
https://arxiv.org/abs/2609.26731
PDF
-
NAWE: Digital Watermarking with Neural-Assisted Watermark Extraction
2026-09-22 10:28:16Roman Chaban, Vitaliy Kinakh, Lilian Rouzaire, Slava VoloshynovskiyAbstract
NAWE (Neural-Assisted Watermark Extraction) combines an explicit signal-processing watermarking construction with a pretrained neural host predictor. A periodic, perceptually masked watermark carrier provides synchronization, Polar coding supplies redundancy, and denoising followed by subtraction extracts the embedded watermark. The denoiser remains frozen, without watermark-specific training. A one-factor-at-a-time study compares Wiener, BM3D, DRUNet, and GS-DRUNet host estimators. Comparisons with TrustMark, SSL Watermarking, PixelSeal, and WAM show NAWE's lowest geometric and photometric class BER and strong message recovery, while filtering and noise remain limitations consistent with the non-adaptive selection of the watermark extractor. The comparison retains the systems' different payloads and coding.
Abstract (translated)
URL
https://arxiv.org/abs/2609.25972
PDF
-
Informed Masking: Structure-Aware Perturbation for Reinforcement Learning in Diffusion Large Language Models
2026-09-22 09:31:23Xiaoyi Yu, Enver Sangineto, Pei Fu, Fiorenzo Parascandolo, Wenhui Tan, Ruikang Zhang, Rita Cucchiara, Ruihua Song, Jian LuanAbstract
Diffusion Large Language Models (dLLMs) have emerged as an efficient alternative to autoregressive models, yet aligning them via Reinforcement Learning (RL) requires likelihood surrogates estimated from masked reconstruction subproblems under a small Monte Carlo budget per rollout. Existing methods construct these subproblems by uniform random masking, leaving open the question of which subproblems to prioritize. We identify a systematic upstream/downstream structure in dLLM rollouts. Some tokens, when revealed, trigger large confidence changes in nearby undecoded positions; we call them upstream. Others induce only small local changes and are therefore downstream. We find masking downstream tokens yields substantially better-posed subproblems than masking upstream tokens, a phenomenon we term subproblem difficulty asymmetry. Based on the observation, we propose Informed Masking (IM), which derives a per-token priority score from the denoising trajectory at zero extra inference cost and biases mask sampling toward downstream tokens. IM is plug-and-play: when plugged into three state-of-the-art dLLM RL methods on LLaDA-8B-Instruct, it delivers up to 2.01%, 8.68%, and 5.77% relative average gains on math and planning benchmarks with improved training stability.
Abstract (translated)
URL
https://arxiv.org/abs/2609.25927
PDF
-
Ultra-fast Neural Inference for Stochastic Gaussian Splatting Denoising
2026-09-22 02:51:29Chenxiao Hu, Hao Zhang, Yanchen Zhang, Meng Gai, Guoping Wang, Sheng LiAbstract
Stochastic rendering eliminates the sorting and alpha blending process in Gaussian splatting, at the cost of introducing spatial noise. Formulating temporal denoising over the pixel stream shared by view-consistent stochastic splatting renderers, we propose a temporal neural denoiser validated on stochastic 2D Gaussian Splatting rendering, combining dual-path exponential moving average accumulation, per-pixel learned trust prediction for history validation, a fixed anisotropic spatial filter and a variance-gated composition with stabilization. The denoiser suppresses the noise, achieving temporally stable, visually compelling outputs during free camera navigation, all while retaining the sort-free, blend-free rasterization performance. The combined pipeline retains a PSNR gap to sorted alpha-blending renderers, but the denoiser's overhead stays below the time saved by removing sorting and blending.
Abstract (translated)
URL
https://arxiv.org/abs/2609.25604
PDF
-
JAMB: Joint Action-Motion Diffusion for Bimanual Manipulation
2026-09-21 19:11:38Chuyang Xiao, Peilin Meng, David HeldAbstract
Coordinated bimanual manipulation is challenging because the motion of either arm can alter the shared 3D scene and thereby affect the other arm. Yet most diffusion policies generate actions without explicitly modeling these future geometric consequences, while predictive variants typically use future state only as auxiliary supervision or fixed conditioning. We address this limitation by proposing JAMB, a diffusion policy that jointly denoises bimanual actions and future 3D point tracks. By allowing action and track hypotheses to evolve together within a shared Transformer, each can inform and refine the other throughout denoising. We further ground multimodal representations in a shared spatiotemporal coordinate system to facilitate geometry-aware interaction during joint denoising. We evaluate JAMB on diverse bimanual manipulation tasks in RoboTwin 2.0 and on a real-world robot, comparing it with action-only policies and alternative future-prediction approaches spanning different state representations and learning objectives. Across 16 simulation tasks, JAMB achieves an average success rate of 83.4%, outperforming the strongest baseline by 23.9 percentage points. On three real-world tasks, it outperforms the action-only and auxiliary geometry prediction methods by 50.0 and 21.2 percentage points, respectively. Beyond these performance gains, JAMB shows stronger generalization to cluttered scenes and out-of-distribution backgrounds than the evaluated baselines. Together, these results demonstrate the effectiveness of our joint action-motion modeling framework for coordinated bimanual manipulation. Our project website is available at this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2609.25322
PDF
