ICLR 2026 论文列表

Video diffusion models (DMs) have enabled high-quality video synthesis. Yet, their substantial computational and memory demands pose serious challenges to real-world deployment, even on high-end GPUs. As a commonly adopted solution, quantization has proven notable success in reducing cost for image DMs, while its direct application to video DMs remains ineffective. In this paper, we present *QVGen*, a novel quantization-aware training (QAT) framework tailored for high-performance and inference-efficient video DMs under extremely low-bit quantization (*e.g.*, $4$-bit or below). We begin with a theoretical analysis demonstrating that reducing the gradient norm is essential to facilitate convergence for QAT. To this end, we introduce auxiliary modules ($\Phi$) to mitigate large quantization errors, leading to significantly enhanced convergence. To eliminate the inference overhead of $\Phi$, we propose a *rank-decay* strategy that progressively eliminates $\Phi$. Specifically, we repeatedly employ singular value decomposition (SVD) and a proposed rank-based regularization $\mathbf{\gamma}$ to identify and decay low-contributing components. This strategy retains performance while zeroing out additional inference overhead. Extensive experiments across $4$ state-of-the-art (SOTA) video DMs, with parameter sizes ranging from $1.3\text{B}\sim14\text{B}$, show that QVGen is *the first* to reach full-precision comparable quality under $4$-bit settings. Moreover, it significantly outperforms existing methods. For instance, our $3$-bit CogVideoX-2B achieves improvements of $+25.28$ in Dynamic Degree and $+8.43$ in Scene Consistency on VBench. Code and models are available at https://github.com/ModelTC/QVGen.

This paper studies the complexity of finding an $\epsilon$-stationary point for stochastic bilevel optimization when the upper-level problem is nonconvex and the lower-level problem is strongly convex. Recent work proposed the first-order method, F${}^2$SA, achieving the $\tilde{\mathcal{O}}(\epsilon^{-6})$ upper complexity bound for first-order smooth problems. This is slower than the optimal $\Omega(\epsilon^{-4})$ complexity lower bound in its single-level counterpart. In this work, we show that faster rates are achievable for higher-order smooth problems. We first reformulate F$^2$SA as approximating the hyper-gradient with a forward difference. Based on this observation, we propose a class of methods F${}^2$SA-$p$ that uses $p$th-order finite difference for hyper-gradient approximation and improves the upper bound to $\tilde{\mathcal{O}}(p \epsilon^{-4-2/p})$ for $p$th-order smooth problems. Finally, we demonstrate that the $\Omega(\epsilon^{-4})$ lower bound also holds for stochastic bilevel problems when the high-order smoothness holds for the lower-level variable, indicating that the upper bound of F${}^2$SA-$p$ is nearly optimal in the highly smooth region $p = \Omega( \log \epsilon^{-1} / \log \log \epsilon^{-1})$.

The scaling of Generative AI (GenAI) models into the hundreds of billions of parameters makes low-precision computation indispensable for efficient deployment. We argue that the fundamental solution lies in developing low-precision \emph{floating-point} formats, which inherently provide numerical stability, memory savings, and hardware efficiency without dequantization overhead. In this paper, we present a theoretical and empirical study of an \emph{exponent concentration} phenomenon in GenAI weights: exponents consistently exhibit low entropy across architectures and modalities. We show that this arises naturally from $\alpha$-stable distributions induced by stochastic gradient descent, and we prove tight bounds on the entropy of exponents. Our analysis establishes a theoretical compression limit near FP4.67, which motivates the design of a practical FP8 format. Building on these insights, we propose \textbf{Exponent-Concentrated FP8 (ECF8)}, a lossless compression framework with entropy-aware encoding and GPU-optimized decoding. Experiments on LLMs and DiTs up to 671B parameters demonstrate up to 26.9\% memory savings and 177.1\% throughput acceleration, with perfectly lossless computations, i.e., no deviation in model outputs. Our results establish exponent concentration as a statistical law of trained models and open a principled path for lossless low-precision floating-point design in the FP8 era.

Neural networks necessitate nonlinearities to achieve universal approximability. Traditional activation functions introduce nonlinearities through rigid feature rectifications. Recent self-gated variants improve traditional methods in fitting flexibility by incorporating learnable content-aware factors and non-local dependencies, enabling dynamic adjustments to activation curves via adaptive translation and scaling. While SOTA approaches achieve notable gains in conventional CNN layers, they struggle to enhance Transformer layers, where fine-grained context is inherently modeled, severely reducing the effectiveness of non-local dependencies leveraged in activation processes. We refer to this critical yet unexplored challenge as the **non-local tension** of activation. Drawing on a decision-making perspective, we systematically analyze the origins of the non-local tension problem and explore the initial solution to foster a more discriminative and generalizable neural activation methodology. This is achieved by rethinking how non-local cues are encoded and transformed into adaptive scaling coefficients, which in turn recalibrate the contributions of features to filter updates through neural activation. Grounded in these insights, we present **FleS**, a novel self-gated activation model for discriminative pattern recognition. Extensive experiments on various popular benchmarks validate our interpretable methodology for improving neural activation modeling.

We initiate the systematic study of decision-theoretic metrics in the design and analysis of algorithms with machine-learned predictions. We introduce approaches based on both deterministic measures such as distance-based evaluation, that help us quantify how close the algorithm is to an ideal solution, and stochastic measures that balance the trade-off between the algorithm's performance and the risk associated with the imperfect oracle. These approaches allow us to quantify the algorithm's performance across the full spectrum of the prediction error, and thus choose the best algorithm within an entire class of otherwise incomparable ones. We apply our framework to three well-known problems from online decision making, namely ski-rental, one-max search, and contract scheduling.

Large Vision-Language Models (VLMs) have achieved remarkable progress in multimodal understanding, yet they struggle when reasoning over information-intensive images that densely interleave textual annotations with fine-grained graphical elements. The main challenges lie in precisely localizing critical cues in dense layouts and multi-hop reasoning to integrate dispersed evidence. We propose Speculative Verdict (SV), a training-free framework inspired by speculative decoding that combines multiple lightweight draft experts with a large verdict model. In the draft stage, small VLMs act as draft experts to generate reasoning paths that provide diverse localization candidates; in the verdict stage, a strong VLM synthesizes these paths to produce the final answer, minimizing computational cost while recovering correct answers. To further improve both efficiency and accuracy, SV introduces a consensus expert selection mechanism that forwards only high-agreement reasoning paths to the verdict. Empirically, SV achieves consistent gains on challenging information-intensive and high-resolution visual question answering benchmarks, including InfographicVQA, ChartMuseum, ChartQAPro, and HR-Bench 4K. By synthesizing correct insights from partially accurate reasoning paths, SV achieves both error correction and cost-efficiency compared to large proprietary models or training pipelines.

The rapid advancement of video generation models has enabled the creation of highly realistic synthetic media, raising significant societal concerns regarding the spread of misinformation. However, current detection methods suffer from critical limitations. They often rely on preprocessing operations like fixed-resolution resizing and cropping, which not only discard subtle, high-frequency forgery artifacts but can also cause distortion and significant information loss. Furthermore, these methods are frequently trained and evaluated on outdated datasets that fail to capture the sophistication of modern generative models. To address these challenges, we introduce two key contributions: a new large-scale dataset and benchmark, as well as a novel detection framework. We present a comprehensive dataset of over 140K videos from 15 state-of-the-art open-source and leading commercial generators. Specifically, we curate the Magic Videos Testset, featuring ultra-realistic videos produced by six latest generators through a meticulous generation and filtering pipeline. In addition, we propose a novel detection framework built on the Qwen2.5-VL Vision Transformer, which processes videos at their native spatial resolution and temporal duration. This native-scale approach preserves high-frequency details and spatiotemporal inconsistencies that are often lost during conventional preprocessing. Extensive experiments show that our method achieves state-of-the-art performance across multiple benchmarks. Our work underscores the importance of native-scale processing and establishes a robust new baseline for AI-generated video detection.

The composition of specialized tools offers a powerful approach for complex visual reasoning, particularly for tasks involving 3D spatial understanding. However, existing visual programming methods are often constrained by fixed toolsets or offline tool induction, which leads to suboptimal solutions and poor tool reuse. We introduce Transductive Visual Programming (TVP), a novel framework that dynamically evolves a library of reusable tools by learning from its problem-solving experience. TVP abstracts recurring solution patterns into new, higher-level tools, which are then used to construct simpler and more effective programs for new tasks. On the challenging Omni3D-Bench, TVP establishes a new state of the art, outperforming both specialized vision-language models and prior visual programming systems. The evolved tools also exhibit strong generalization to out-of-domain queries on 3DSRBench, SpatialSense, and VGBench. Our work demonstrates that transductive tool evolution is a powerful and generalizable paradigm for building robust visual reasoning systems.

The dueling bandits (DB) problem addresses online learning from relative preferences, where the learner queries pairs of arms and receives binary win-loss feedback. Most existing work focuses on designing algorithms for specific stochastic or adversarial environments. Recently, a unified algorithm has been proposed that achieves convergence across all settings. However, this approach relies on the existence of a Condorcet winner, which is often not achievable, particularly when the preference matrix changes in the adversarial setting. Aiming for a more general Borda winner objective, there currently exists no unified framework that simultaneously achieves optimal regret across these environments. In this paper, we explore how the follow-the-regularized-leader (FTRL) algorithm can be employed to achieve this objective. We propose a hybrid negative entropy regularizer and demonstrate that it enables us to achieve $\tilde{O}(K^{1/3} T^{2/3})$ regret in the adversarial setting, ${O}({K \log^2 T}/{\Delta_{\min}^2})$ regret in the stochastic setting, and $O({K \log^2 T }/{\Delta_{\min}^2} + ({C^2 K \log^2 T }/{\Delta_{\min}^2})^{1/3})$ regret in the corrupted setting, where $K$ is the arm set size, $T$ is the horizon, $\Delta_{\min}$ is the minimum gap between the optimal and sub-optimal arms, and $C$ is the corruption level. These results align with the state-of-the-art in individual settings, while eliminating the need to assume a specific environment type. We also present experimental results demonstrating the advantages of our algorithm over baseline methods across different environments.

We propose Neogradon, a video DiT (Diffusion Transformer) designed to run on a low-power NPU present in devices such as phones and laptop computers. We demonstrate that, despite video transformers' huge memory and compute cost, mobile devices can run these models when carefully optimised for efficiency. To achieve this level of efficiency, i) we replace the original large Text-Encoder with a much smaller one with minimal quality loss through our novel distillation framework which doesn’t require any image or video data. ii) We propose an Asymmetric Decoder distillation approach which allows us to replace the native codec-latent-VAE decoder with a more efficient one, without disturbing the generative latent-space of the video generation pipeline. iii) With our Block Pruning strategy, we remove entire blocks from the MMDiT denoiser based on their relative importance and recover original performance through a two-stage distillation process. iv) We reduce the diffusion sampling cost using our novel extended version of DMD (Distribution Matching Distillation) for the Pyramidal Flow-Matching objective. Neodragon generates 49 frames of [640$\times$1024] resolution within 7.6 seconds on the Qualcomm Hexagon NPU with a VBench total score of 81.61, setting a new state-of-the-art for mobile video generation.

In this work, we present Transitive Reinforcement Learning (TRL), a new value learning algorithm based on a divide-and-conquer paradigm. TRL is designed for offline goal-conditioned reinforcement learning (GCRL) problems, where the aim is to find a policy that can reach any state from any other state in the smallest number of steps. TRL converts a triangle inequality structure present in GCRL into a practical divide-and-conquer value update rule. This has several advantages compared to alternative value learning paradigms. Compared to temporal difference (TD) methods, TRL suffers less from bias accumulation, as in principle it only requires $O(\log T)$ recursions (as opposed to $O(T)$ in TD learning) to handle a length-T trajectory. Unlike Monte Carlo methods, TRL suffers less from high variance as it performs dynamic programming. Experimentally, we show that TRL achieves the best performance in highly challenging, long-horizon benchmark tasks compared to previous offline GCRL algorithms.

Video matting is crucial for applications such as film production and virtual reality, yet deploying its computationally intensive models on resource-constrained devices presents challenges. Quantization is a key technique for model compression and acceleration. As an efficient approach, Post-Training Quantization (PTQ) is still in its nascent stages for video matting, facing significant hurdles in maintaining accuracy and temporal coherence. To address these challenges, this paper proposes a novel and general PTQ framework specifically designed for video matting models, marking, to the best of our knowledge, the first systematic attempt in this domain. Our contributions include: (1) A two-stage PTQ strategy that combines block reconstruction-based optimization for fast, stable initial quantization and local dependency capture, followed by a global calibration of quantization parameters to minimize accuracy loss. (2) A Statistically-Driven Global Affine Calibration (GAC) method that enables the network to compensate for cumulative statistical distortions arising from factors such as neglected BN layer effects, even reducing the error of existing PTQ methods on video matting tasks up to 20%. (3) An Optical Flow Assistance (OFA) component that leverages temporal and semantic priors from frames to guide the PTQ process, enhancing the model’s ability to distinguish moving foregrounds in complex scenes and ultimately achieving near full-precision performance even under ultra-low-bit quantization. Comprehensive quantitative and visual results show that our PTQ4VM achieves the state-of-the-art accuracy performance across different bit-widths compared to the existing quantization methods. We highlight that the 4-bit PTQ4VM even achieves performance close to the full-precision counterpart while enjoying 8× FLOP savings.

Post-hoc instance-level graph neural network (GNN) explainers are developed to identify a compact subgraph (i.e., explanation) that encompasses the most influential components for each input graph. A fundamental limitation of existing methods lies in the insufficient utilization of structural information during GNN explainer optimization. They typically optimize the explainer by aligning the GNN predictions of input graph and its explanation in the graph label space which inherently lacks expressiveness to describe various graph structures. Motivated by the powerful structural expression ability of vectorized graph representations, we for the first time propose to shift the GNN explainer optimization from the graph label space to the graph representation space. However, the paradigm shift is challenging due to both the entanglement between the explanatory and non-explanatory substructures, and the distributional discrepancy between the input graph and the explanation subgraph. To this end, we meticulously design IDEA, a universal dual-stage optimization framework grounded in a prototypical graph representation space, which can generalize across diverse existing GNN explainer architectures. Specifically, in the Structural Information Disentanglement stage, a graph tokenizer equipped with a structure-aware disentanglement objective is designed to disentangle the explanatory substructures and encapsulate them into explanatory prototypes. In the Explanatory Prototype Alignment stage, IDEA aligns the representational distributions of the input graph and its explanation unified in the prototypical representation space, to optimize the GNN explainer. Comprehensive experiments on real-world and synthetic datasets demonstrate the effectiveness of IDEA, with the average improvements of ROC-AUC by 4.45% and precision by 48.71%. We further integrate IDEA with diverse explainer architectures and achieve an improvement by up to 10.70%, which verifies its generalizability.

Large-scale pretrained diffusion models have significantly enhanced the quality of generated videos, and yet their use in real-time streaming remains limited. Autoregressive models offer a natural framework for sequential frame synthesis but require heavy computation to achieve high fidelity. Diffusion distillation can compress these models into efficient few-step variants, but existing video distillation approaches largely adapt image-specific methods that neglect temporal dependencies. These techniques often excel in image generation but underperform in video synthesis, exhibiting reduced motion coherence, error accumulation over long sequences, and a latency–quality trade-off. We identify two factors that result in these limitations: insufficient utilization of temporal context during step reduction and implicit prediction of subsequent noise levels in next-chunk prediction (exposure bias). To address these issues, we propose Diagonal Distillation, which operates orthogonally to existing approaches and better exploits temporal information across both video chunks and denoising steps. Central to our approach is an asymmetric generation strategy: more steps early, fewer steps later. This design allows later chunks to inherit rich appearance information from thoroughly processed early chunks, while using partially denoised chunks as conditional inputs for subsequent synthesis. By aligning the implicit prediction of subsequent noise levels during chunk generation with the actual inference conditions, our approach mitigates error propagation and reduces oversaturation in long-range sequences. We further incorporate implicit optical flow modeling to preserve motion quality under strict step constraints. Our method generates a 5-second video in just 2.61 seconds (up to 31 FPS), achieving a 277.3× speedup over the undistilled model and doubling the acceleration ratio of the state-of-the-art (140×) without sacrificing visual quality. The source code and trained models will be released and we have uploaded the video to an anonymous hosting service at: https://diagonal-distillation.github.io/.

Diffusion models have achieved remarkable success in generative modeling. Despite more stable training, the loss of diffusion models is not indicative of absolute data-fitting quality, since its optimal value is typically not zero but unknown, leading to the confusion between large optimal loss and insufficient model capacity. In this work, we advocate the need to estimate the optimal loss value for diagnosing and improving diffusion models. We first derive the optimal loss in closed form under a unified formulation of diffusion models, and develop effective estimators for it, including a stochastic variant scalable to large datasets with proper control of variance and bias. With this tool, we unlock the inherent metric for diagnosing training quality of representative diffusion model variants, and develop a more performant training schedule based on the optimal loss. Moreover, using models with 120M to 1.5B parameters, we find that the power law is better demonstrated after subtracting the optimal loss from the actual training loss, suggesting a more principled setting for investigating the scaling law for diffusion models.

The objective of machine unlearning (MU) is to eliminate previously learned data from a model. However, it can be challenging to strike a balance between computation cost and performance when using existing MU techniques. Taking inspiration from the influence of label smoothing on model confidence and differential privacy, we propose a simple gradient-based MU approach that uses an inverse process of label smoothing. This work introduces UGradSL, a simple, plug-and-play MU approach that uses smoothed labels. We provide theoretical analyses demonstrating why properly introducing label smoothing improves MU performance. We conducted extensive experiments on several datasets of various sizes and different modalities, demonstrating the effectiveness and robustness of our proposed method. UGradSL also shows close connection to improve the local differential privacy. The consistent improvement in MU performance is only at a marginal cost of additional computations. For instance, UGradSL improves over the gradient ascent MU baseline constantly on different unlearning tasks without sacrificing unlearning efficiency. A self-adaptive UGradSL is also given for simple parameter selection.

Transformers robustly exhibit the ability to perform in-context learning, whereby their predictive accuracy on a task can increase not by parameter updates but merely with the placement of training samples in their context windows. Recent works have shown that transformers achieve this by implementing gradient descent in their forward passes. Such results, however, are restricted to standard transformer architectures, which handle finite-dimensional inputs. In the space of PDE surrogate modeling, a generalization of transformers to handle infinite-dimensional function inputs, known as "continuum transformers," has been proposed and similarly observed to exhibit in-context learning. Despite impressive empirical performance, such in-context learning has yet to be theoretically characterized. We herein demonstrate that continuum transformers perform in-context operator learning by performing gradient descent in an operator RKHS. We demonstrate this using novel proof strategies that leverage a generalized representer theorem for Hilbert spaces and gradient flows over the space of functionals of a Hilbert space. We additionally show the operator learned in context is the Bayes Optimal Predictor in the infinite depth limit of the transformer. We then provide empirical validations of this optimality result and demonstrate that the parameters under which such gradient descent is performed are recovered through the continuum transformer training.

AI agents are increasingly being deployed to automate tasks, often based on underspecified user instructions. Making unwarranted assumptions to compensate for the missing information and failing to ask clarifying questions can lead to suboptimal outcomes, safety risks due to tool misuse, and wasted computational resources. In this work, we study the ability of LLM agents to handle underspecified instructions in interactive code generation settings by evaluating proprietary and open-weight models on their performance across three key steps: (a) detecting underspecificity, (b) asking targeted clarification questions, and (c) leveraging the interaction to improve performance in underspecified scenarios. Our findings reveal that models struggle to distinguish between well-specified and underspecified instructions. However, when models interact for underspecified inputs, they effectively obtain vital information from the user leading to significant improvements in performance, up to 74\% over the non-interactive settings, underscoring the value of effective interaction. Our study highlights critical gaps in how current state-of-the-art models handle missing information in complex software engineering tasks and structures the evaluation into distinct steps to enable targeted improvements.

Protein language models (PLMs) have enabled advances in structure prediction and de novo protein design, yet they frequently collapse into pathological repetition during generation. Unlike in text, where repetition merely reduces readability, in proteins it undermines structural confidence and functional viability. To unify this problem, we present the first systematic study of repetition in PLMs. We first propose quantitative metrics to characterize motif-level and homopolymer repetition and then demonstrate their negative impact on folding reliability. To address this challenge, we propose UCCS (Utility-Controlled Contrastive Steering), which steers protein generation with a constrained dataset. Instead of naively contrasting high- vs. low-repetition sequences, we construct contrastive sets that maximize differences in repetition while tightly controlling for structural utility. This disentanglement yields steering vectors that specifically target repetition without degrading foldability. Injected at inference, these vectors consistently reduce repetition without retraining or heuristic decoding. Experiments with ESM-3 and ProtGPT2 in CATH, UniRef50, and SCOP show that our method outperforms decoding penalties and other baselines, substantially lowering repetition while preserving AlphaFold confidence scores. Our results establish repetition control as a central challenge for PLMs and highlight dataset-guided steering as a principled approach for reliable protein generation.