ICLR 2026 论文列表

Mixture-of-Experts (MoE) enables efficient scaling of large language models by activating only a subset of experts per input token. However, deploying MoE-based models incurs significant memory overhead due to the need to retain all experts in memory. While structured pruning is promising to reduce memory costs, existing methods often show suboptimal performance and unstable degradation in three dimensions: model architectures, calibration data sources, and calibration sample sizes. This paper proposes \textbf{M}ixture-\textbf{o}f-\textbf{N}ovices-and-\textbf{E}xperts (\textbf{MoNE}), a novel expert pruning method that replaces redundant experts with lightweight novices to achieve effective and robust model compression. MoNE evaluates expert redundancy based on two metrics: access frequency and output variance. Experts exhibiting low usage and stable outputs are pruned and replaced with lightweight novices—unbiased estimations of their original outputs—minimizing performance degradation. Extensive experiments demonstrate that MoNE consistently outperforms baseline methods with minimal accuracy degradation across the three dimensions, confirming its effectiveness and robustness. Notably, it outperforms baselines by up to 2.72 for the average zero shot accuracy across nine downstream tasks under 25\% pruning ratio, with only 0.14 performance drop for Qwen2-57B-A14B. The code is available at \url{https://github.com/zxgx/mode-pd}.

Mixture-of-Experts (MoE) has become a dominant architecture for scaling Large Language Models (LLMs) efficiently by decoupling total parameters from computational cost. However, this decoupling creates a critical challenge: predicting the model capacity of a given MoE configurations (e.g., expert activation ratio and granularity) remains an unresolved problem. To address this gap, we introduce Efficiency Leverage (EL), a metric quantifying the computational advantage of an MoE model over a dense equivalent. We conduct a large-scale empirical study, training over 300 models up to 28B parameters, to systematically investigate the relationship between MoE architectural configurations and EL. Our findings reveal that EL is primarily driven by the expert activation ratio and the total compute budget, both following predictable power laws, while expert granularity acts as a non-linear modulator with a clear optimal range. We integrate these discoveries into a unified scaling law that accurately predicts the EL of an MoE architecture based on its configuration. To validate our derived scaling laws, we designed and trained MoE-mini, a model with only 0.85B active parameters, alongside a 6.1B dense model for comparison. When trained on an identical 1T high-quality token dataset, MoE-mini matched the performance of the 6.1B dense model while consuming over 7x fewer computational resources, thereby confirming the accuracy of our scaling laws. This work provides a principled and empirically-grounded foundation for the scaling of efficient MoE models.

Autoregressive large language models (LLMs) have achieved remarkable improvement across many tasks but incur high computational and memory costs. Knowledge distillation (KD) mitigates this issue by transferring knowledge from a large teacher to a smaller student through distributional alignment. Previous studies have proposed various discrepancy metrics, but the capacity gap and training instability caused by near-zero probabilities, stemming from the high-dimensional output of LLMs, remain fundamental limitations. To overcome these challenges, several approaches implicitly or explicitly incorporating assistant distribution have recently been proposed. However, the past proposals of assistant distributions have been a fragmented approach without a systematic investigation of the interpolation path and the divergence. This paper proposes $\alpha$-mixture assistant distribution, a novel generalized family of assistant distributions, and $\alpha$-mixture distillation, coined AMiD, a unified framework for KD using the assistant distribution. The $\alpha$-mixture assistant distribution provides a continuous extension of the assistant distribution by introducing a new distribution design variable $\alpha$, which has been fixed in all previous approaches. Furthermore, AMiD generalizes the family of divergences used with the assistant distributions based on optimality, which has also been restricted in previous works. Through extensive experiments, we demonstrate that AMiD offers superior performance and training stability by leveraging a broader and theoretically grounded assistant distribution space.

While large and diverse datasets have driven recent advances in large models, identifying the optimal data mixture for pre-training and post-training remains a significant open problem. We address this challenge with FastMix, a novel framework that automates data mixture discovery while training only a single proxy model. Instead of relying on predefined heuristics or resource-intensive simulations, FastMix jointly optimizes mixture coefficients and model parameters, substantially improving efficiency and scalability over prior approaches. At the core of FastMix is a reformulation of mixture selection as a bilevel optimization problem. Under this reformulation, we show that optimizing mixture ratios is mathematically equivalent to assigning per-source loss weights under uniform source sampling. This embeds the mixture coefficients directly into the differentiable iterative optimization objective, enabling efficient, gradient-based optimization of both mixture and model. To solve the optimization problem, FastMix implements an approximate iterative optimization procedure, alternating between (i) updating model parameters on data sampled according to current mixture ratios (inner loop) and (ii) updating mixture ratios based on validation feedback (outer loop). Across pre- and post-training, FastMix outperforms baselines while drastically reducing search cost: in pre-training, it attains an average score of 48.2 with 1.3 GPU-hours ($\times 550$ vs. RegMix; $\times 55$ vs. CLIMB), and in post-training (SFT) it leads with 65.4 with a $+5.5$ gain over the next best, completing search in 2.2 GPU-hours compared to the 115 GPU-hours required by CLIMB/RegMix.

Current state-of-the-art methods for out-of-distribution (OOD) generalization lack the ability to effectively address datasets with heterogeneous causal subgraphs at the instance level. Existing approaches that attempt to handle such heterogeneity either rely on data augmentation, which risks altering label semantics, or impose causal assumptions whose validity in real-world datasets is uncertain. We introduce a novel *Mixture-of-Experts (MoE)* framework that can model heterogeneous causal subgraphs without relying on restrictive assumptions. Our key idea is to address instance-level heterogeneity by enforcing semantic *diversity* among experts, each generating a distinct causal subgraph, while a learned gate assigns *sparse* weights that adaptively focus on the most relevant experts for each input. Our theoretical analysis shows that these two properties jointly reduce OOD error. In practice, our experts are scalable and do not require environment labels. Empirically, our framework achieves strong performance on the GOOD benchmark across both synthetic and real-world structural shifts.

Recent studies have shown that combining parameter-efficient fine-tuning (PEFT) with mixture-of-experts (MoE) is an effective strategy for adapting large language models (LLMs) to the downstream tasks. However, most existing approaches rely on conventional TopK routing, which requires careful hyperparameter tuning and assigns a fixed number of experts to each token. In this work, we propose LD-MoLE, a Learnable Dynamic routing mechanism for Mixture of LoRA Experts that enables adaptive, token-dependent, and layer-wise expert allocation. Our method replaces the non-differentiable TopK selection with a differentiable routing function and a closed-form solution. Moreover, our design allows the model to adaptively determine the number of experts to activate for each token at different layers. In addition, we introduce an analytical sparsity control objective to regularize the number of activated experts. Extensive experiments on the Qwen3-1.7B and Llama-3.2-3B models show that LD-MoLE achieves the highest average scores compared to state-of-the-art baselines, across a diverse set of benchmarks. Our method not only achieves superior performance, but also demonstrates the ability to learn token-dependent and layer-wise expert allocation.

Sparse Mixture-of-Experts (MoE) have been widely adopted in recent large language models since it can efficiently scale up the model capability without increasing the inference cost. However, evaluations on broad downstream tasks reveal a consistent suboptimality of the routers in existing MoE LLMs, which results in a severe performance gap (e.g., 10-20% in accuracy) to the optimal routing. In this paper, we show that aligning the manifold of routing weights with that of task embedding via post-training can effectively reduce the gap and improve MoE LLMs’ generalization performance. Our method, “Routing Manifold Alignment (RoMA)”, introduces an additional manifold regularization term in the post-training objective and only requires lightweight finetuning of routers (with other parameters frozen). Specifically, the regularization encourages the routing weights of each sample to be close to those of its successful neighbors (whose routing weights lead to correct answers) in a task embedding space. Consequently, samples targeting similar tasks will share similar expert choices across layers. Building such bindings between tasks and experts over different samples is essential to achieve better generalization. Moreover, RoMA demonstrates the advantage of unifying the task understanding (by embedding models) with solution generation (by MoE LLMs). In experiments, we finetune routers in two recent MoE LLMs using RoMA. Evaluations on diverse benchmarks and extensive comparisons with baselines show the substantial improvement brought by RoMA.

While massively both scaling data and models have become central in NLP and 2D vision, their benefits for 3D point cloud understanding remain limited. We study the initial step of 3D point cloud scaling under a realistic regime: large-scale multi-dataset joint training for 3D semantic segmentation, with no dataset labels available at inference time. Point clouds arise from a wide range of sensors (e.g., depth cameras, LiDAR) and scenes (e.g., indoor, outdoor), yielding heterogeneous scanning patterns, sampling densities, and semantic biases; naively mixing such datasets degrades standard backbones. We introduce **Point-MoE**, a Mixture-of-Experts design that expands capacity through sparsely activated expert MLPs and a lightweight top-$k$ router, allowing tokens to select specialized experts without requiring dataset supervision. Trained jointly on a diverse mix of indoor and outdoor datasets and evaluated on seen datasets and in zero-shot settings, Point-MoE outperforms prior methods without using dataset labels for either training or inference. This outlines a scalable path for 3D perception: letting the model discover structure in heterogeneous 3D data rather than imposing it via manual curation or dataset-specific heuristics.

Mixture-of-Experts (MoE) enables efficient scaling of large language models (LLMs) with sparsely activated experts during inference. To effectively deploy large MoE models on memory-constrained devices, many systems introduce expert offloading which caches a subset of experts in fast memory, leaving others on slow memory to run on CPU or load on demand. While some research has exploited the locality of expert activations, where consecutive tokens activate similar experts, the degree of this **local routing consistency** varies across models and remains understudied. In this paper, we propose two metrics to measure local routing consistency of MoE models: (1) **Segment Routing best Performance (SRP)**, which evaluates how well a fixed group of experts can cover the needs of a segment of tokens, and (2) **Segment Cache best Hit rate (SCH)**, which measures the hit rate of an expert cache utilizing a length of future information under a cache limit. We analyze 20 MoE LLMs with diverse sizes and architectures and use toy models to verify key factors related to local routing consistency. We find a strong trade-off between local routing consistency and *local* load balance, while showing that *global* load balance can coexist with local routing consistency. Meanwhile, settings like shared experts that decrease expert combination space can lead to low local routing consistency. We further reveal that domain-specialized experts contribute more to routing consistency than vocabulary-specialized ones, and that most models balance between cache effectiveness and efficiency with cache sizes approximately twice the active experts. These findings pave the way for memory-efficient MoE design and deployment without compromising inference speed.

Mixture-of-Experts (MoE) has emerged as a powerful paradigm for scaling model capacity while preserving computational efficiency. Despite its notable success in large language models (LLMs), existing attempts to apply MoE to Diffusion Transformers (DiTs) have yielded limited gains. We attribute this gap to fundamental differences between language and visual tokens. Language tokens are semantically dense with pronounced inter-token variation, while visual tokens exhibit spatial redundancy and functional heterogeneity, hindering expert specialization in vision MoE. To this end, we present $\textbf{ProMoE}$, an MoE framework featuring a two-step router with explicit routing guidance that promotes expert specialization. Specifically, this guidance encourages the router to $\textit{first}$ partition image tokens into conditional and unconditional sets via conditional routing according to their functional roles, and $\textit{second}$ refine the assignments of conditional image tokens through prototypical routing with learnable prototypes based on semantic content. Moreover, the similarity-based expert allocation in latent space enabled by prototypical routing offers a natural mechanism for incorporating explicit semantic guidance, and we validate that such guidance is crucial for vision MoE. Building on this, we propose a routing contrastive loss that explicitly enhances the prototypical routing process, promoting intra-expert coherence and inter-expert diversity. Extensive experiments on ImageNet benchmark demonstrate that ProMoE surpasses state-of-the-art methods under both Rectified Flow and DDPM training objectives. Code and models will be made publicly available.

Generative models can generate photorealistic images at scale. This raises urgent concerns about the ability to detect synthetically generated images and attribute these images to specific sources. While watermarking has emerged as a possible solution, existing methods remain fragile to realistic distortions, susceptible to adaptive removal, and expensive to update when the underlying watermarking key changes. We propose a general watermarking framework that formulates the encoding problem as key-dependent perturbation of the parameters of a generative model. Within this framework, we introduce Mixture of LoRA Markers (MOLM), a routing-based instantiation in which binary keys activate lightweight LoRA adapters inside residual and attention blocks. This design avoids key-specific re-training and achieves the desired properties such as imperceptibility, fidelity, verifiability, and robustness. Experiments on Stable Diffusion and FLUX show that MOLM preserves image quality while achieving robust key recovery against distortions, compression and regeneration, averaging attacks, and black-box adversarial attacks on the extractor.

Long video generation with Diffusion Transformers (DiTs) is bottlenecked by the quadratic scaling of full attention with sequence length. Since attention is highly redundant, outputs are dominated by a small subset of query–key pairs. Existing sparse methods rely on blockwise coarse estimation, whose accuracy–efficiency trade-offs are constrained by block size. This paper introduces Mixture-of-Groups Attention (MoGA), an efficient sparse attention that uses a lightweight, learnable token router to precisely match tokens without blockwise estimation. Through semantic-aware routing, MoGA enables effective long-range interactions. As a kernel-free method, MoGA integrates seamlessly with modern attention stacks, including FlashAttention and sequence parallelism. Building on MoGA, we develop an efficient long video generation model that end-to-end produces minute-level, multi-shot, 480p videos at 24 fps, with a context length of approximately 580k. Comprehensive experiments on various video generation tasks validate the effectiveness of our approach. We provide an anonymous link \url{https://anonymous.4open.science/r/MoGA} to showcase the generated videos.