ICLR 2025 论文列表

Transformers are widely used as generic backbones in computer vision, despite initially introduced for natural language processing. Recently, the Long Short-Term Memory (LSTM) has been extended to a scalable and performant architecture - the xLSTM - which overcomes long-standing LSTM limitations via exponential gating and parallelizable matrix memory structure. In this paper, we introduce Vision-LSTM (ViL), an adaption of the xLSTM building blocks to computer vision. ViL comprises a stack of xLSTM blocks where odd blocks process the sequence of patch tokens from top to bottom while even blocks go from bottom to top. ViL achieves strong performances on classification, transfer learning and segmentation tasks as well as a beneficial pre-training cost-to-performance trade-off. Experiments show that ViL holds promise to be further deployed as new generic backbone for computer vision architectures.

Generative Foundation Models (GFMs) have achieved remarkable success in producing high-quality synthetic data for images and text. However, their application to tabular data presents significant challenges due to the heterogeneous nature of table features. Current cross-table learning frameworks struggle because they lack a generative model backbone and an effective mechanism to decode heterogeneous feature values. To address these challenges, we propose the Cross-Table Synthesizer (CTSyn), a diffusion-based generative foundation model for tabular data generation. CTSyn comprises two key components. The first is an autoencoder network that consolidates diverse tables into a unified latent space. It dynamically reconstructs table values using a table schema embedding, allowing adaptation to heterogeneous datasets. The second is a conditional latent diffusion model that generates samples from the learned latent space, conditioned on the table schema. Through large-scale pre-training, CTSyn outperforms existing table synthesizers on standard benchmarks in both utility and diversity. These results position CTSyn as a promising framework for synthetic table generation and lay the groundwork for developing large-scale tabular foundation models.

We empirically investigate the camera bias of person re-identification (ReID) models. Previously, camera-aware methods have been proposed to address this issue, but they are largely confined to training domains of the models. We measure the camera bias of ReID models on unseen domains and reveal that camera bias becomes more pronounced under data distribution shifts. As a debiasing method for unseen domain data, we revisit feature normalization on embedding vectors. While the normalization has been used as a straightforward solution, its underlying causes and broader applicability remain unexplored. We analyze why this simple method is effective at reducing bias and show that it can be applied to detailed bias factors such as low-level image properties and body angle. Furthermore, we validate its generalizability across various models and benchmarks, highlighting its potential as a simple yet effective test-time postprocessing method for ReID. In addition, we explore the inherent risk of camera bias in unsupervised learning of ReID models. The unsupervised models remain highly biased towards camera labels even for seen domain data, indicating substantial room for improvement. Based on observations of the negative impact of camera-biased pseudo labels on training, we suggest simple training strategies to mitigate the bias. By applying these strategies to existing unsupervised learning algorithms, we show that significant performance improvements can be achieved with minor modifications.

Spiking neural networks (SNNs) are posited as a computationally efficient and biologically plausible alternative to conventional neural architectures, with their core computational framework primarily using the leaky integrate-and-fire (LIF) neuron model. However, the limited hidden state representation of LIF neurons, characterized by a scalar membrane potential, and sequential spike generation process, poses challenges for effectively developing scalable spiking models to address long-range dependencies in sequence learning tasks. In this study, we develop a scalable probabilistic spiking learning framework for long-range dependency tasks leveraging the fundamentals of state space models. Unlike LIF neurons that rely on the deterministic Heaviside function for a sequential process of spike generation, we introduce a SpikeSampler layer that samples spikes stochastically based on an SSM-based neuronal model while allowing parallel computations. To address non-differentiability of the spiking operation and enable effective training, we also propose a surrogate function tailored for the stochastic nature of the SpikeSampler layer. To enhance inter-neuron communication, we introduce the SpikeMixer block, which integrates spikes from neuron populations in each layer. This is followed by a ClampFuse layer, incorporating a residual connection to capture complex dependencies, enabling scalability of the model. Our models attain state-of-the-art performance among SNN models across diverse long-range dependency tasks, encompassing the Long Range Arena benchmark, permuted sequential MNIST, and the Speech Command dataset and demonstrate sparse spiking pattern highlighting its computational efficiency.

Evaluating the reasoning abilities of large language models (LLMs) is challenging. Existing benchmarks often depend on static datasets, which are vulnerable to data contamination and may get saturated over time, or on binary live human feedback that conflates reasoning with other abilities. As the most prominent dynamic benchmark, Chatbot Arena evaluates open-ended questions in real-world settings, but lacks the granularity in assessing specific reasoning capabilities. We introduce GameArena, a dynamic benchmark designed to evaluate LLM reasoning capabilities through interactive gameplay with humans. GameArena consists of three games designed to test specific reasoning capabilities (e.g., deductive and inductive reasoning), while keeping participants entertained and engaged. We analyze the gaming data retrospectively to uncover the underlying reasoning processes of LLMs and measure their fine-grained reasoning capabilities. We collect over 2000 game sessions and provide detailed assessments of various reasoning capabilities for five state-of-the-art LLMs. Our user study with 100 participants suggests that GameArena improves user engagement compared to Chatbot Arena. For the first time, GameArena enables the collection of step-by-step LLM reasoning data in the wild.

Verifiable formal languages like Lean have profoundly impacted mathematical reasoning, particularly through the use of large language models (LLMs) for automated reasoning. A significant challenge in training LLMs for these formal languages is the lack of parallel datasets that align natural language with formal language proofs. To address this challenge, this paper introduces a novel framework for translating the Mathlib4 corpus (a unified library of mathematics in formal language Lean 4) into natural language. Building upon this, we employ a dual augmentation strategy that combines tactic-based and informal-based approaches, leveraging the Lean-jixia system, a Lean 4 analyzer. We present the results of this pipeline on Mathlib4 as Herald (Hierarchy and Retrieval-based Translated Lean Dataset). We also propose the Herald Translator, which is fine-tuned on Herald. Herald translator achieves a 96.7\% accuracy (Pass@128) on formalizing statements in the miniF2F-test and a 23.5\% accuracy on our internal graduate-level textbook dataset, outperforming InternLM2-Math-Plus-7B (73.0\% and 7.5\%) and TheoremLlama (50.1\% and 4.0\%). Furthermore, we propose a section-level translation framework for real-world applications. As a direct application of Herald translator, we have successfully translated a template section in the Stack project, marking a notable progress in the automatic formalization of graduate-level mathematical literature. Our model, along with the datasets, are open-sourced to the public.

Deep learning architectures for supervised learning on tabular data range from simple multilayer perceptrons (MLP) to sophisticated Transformers and retrieval-augmented methods. This study highlights a major, yet so far overlooked opportunity for substantially improving tabular MLPs; namely, parameter-efficient ensembling -- a paradigm for imitating an ensemble of models with just one model. We start by describing TabM -- a simple model based on MLP and BatchEnsemble (an existing technique), improved with our custom modifications. Then, we perform a large scale evaluation of tabular DL architectures on public benchmarks in terms of both task performance and efficiency, which renders the landscape of tabular DL in a new light. In particular, we find that TabM outperforms prior tabular DL models, while the complexity of attention- and retrieval-based methods does not pay off. Lastly, we conduct a detailed empirical analysis, that sheds some light on the high performance of TabM. For example, we show that parameter-efficient ensembling is not an arbitrary trick, but rather a highly effective way to reduce overfitting and improve optimization dynamics of tabular MLPs. Overall, our work brings an impactful technique to tabular DL, analyses its behaviour, and advances the performance-efficiency tradeoff with TabM -- a simple and powerful baseline for researchers and practitioners.

We revisit the ``dataset classification'' experiment suggested by Torralba & Efros (2011) a decade ago, in the new era with large-scale, diverse, and hopefully less biased datasets as well as more capable neural network architectures. Surprisingly, we observe that modern neural networks can achieve excellent accuracy in classifying which dataset an image is from: e.g., we report 84.7% accuracy on held-out validation data for the three-way classification problem consisting of the YFCC, CC, and DataComp datasets. Our further experiments show that such a dataset classifier could learn semantic features that are generalizable and transferable, which cannot be explained by memorization. We hope our discovery will inspire the community to rethink issues involving dataset bias.

Recently, many self-supervised learning methods for image reconstruction have been proposed that can learn from noisy data alone, bypassing the need for ground-truth references. Most existing methods cluster around two classes: i) Stein's Unbiased Risk Estimate (SURE) and similar approaches that assume full knowledge of the noise distribution, and ii) Noise2Self and similar cross-validation methods that require very mild knowledge about the noise distribution. The first class of methods tends to be impractical, as the noise level is often unknown in real-world applications, and the second class is often suboptimal compared to supervised learning. In this paper, we provide a theoretical framework that characterizes this expressivity-robustness trade-off and propose a new approach based on SURE, but unlike the standard SURE, does not require knowledge about the noise level. Throughout a series of experiments, we show that the proposed estimator outperforms other existing self-supervised methods on various imaging inverse problems.

Recent advancements in large language models (LLMs) reveal a perplexing phenomenon in continual learning: despite extensive training, models experience significant performance declines, raising questions about task alignment and underlying knowledge retention. This study first explores the concept of "spurious forgetting", proposing that such performance drops often reflect a decline in task alignment rather than true knowledge loss. Through controlled experiments with a synthesized dataset, we investigate the dynamics of model performance during the initial training phases of new tasks, discovering that early optimization steps can disrupt previously established task alignments. Our theoretical analysis connects these shifts to orthogonal updates in model weights, providing a robust framework for understanding this behavior. Ultimately, we introduce a Freezing strategy that fix the bottom layers of the model, leading to substantial improvements in four continual learning scenarios. Our findings underscore the critical distinction between task alignment and knowledge retention, paving the way for more effective strategies in continual learning.

Prompt Tuning adapts frozen models to new tasks by prepending a few learnable embeddings to the input. However, it struggles with tasks that suffer from data scarcity. To address this, we explore Cross-Modality Prompt Transfer, leveraging prompts pretrained on a data-rich modality to improve performance on data-scarce tasks in another modality. As a pioneering study, we first verify the feasibility of cross-modality prompt transfer by directly applying frozen source prompts (trained on the source modality) to the target modality task. To empirically study cross-modality prompt transferability, we train a linear layer to adapt source prompts to the target modality, thereby boosting performance and providing ground-truth transfer results. Regarding estimating prompt transferability, existing methods show ineffectiveness in cross-modality scenarios where the gap between source and target tasks is larger. We address this by decomposing the gap into the modality gap and the task gap, which we measure separately to autonomously select the best source prompt for a target task. Additionally, we propose Attention Transfer to further reduce the gaps by injecting target knowledge into the prompt and reorganizing a top-transferable source prompt using an attention block. We conduct extensive experiments involving prompt transfer from 13 source language tasks to 19 target vision tasks under three settings. Our findings demonstrate that: (i) cross-modality prompt transfer is feasible, supported by in-depth analysis; (ii) measuring both the modality and task gaps is crucial for accurate prompt transferability estimation, a factor overlooked by previous studies; (iii) cross-modality prompt transfer can significantly release the powers of prompt tuning on data-scarce tasks, as evidenced by comparisons with a newly released prompt-based benchmark.

Investigating noise distributions beyond Gaussian in diffusion generative models remains an open challenge. The Gaussian case has been a large success experimentally and theoretically, admitting a unified stochastic differential equation (SDE) framework, encompassing score-based and denoising formulations. Recent studies have investigated the potential of \emph{heavy-tailed} noise distributions to mitigate mode collapse and effectively manage datasets exhibiting class imbalance, heavy tails, or prominent outliers. Very recently, Yoon et al.\ (NeurIPS 2023), presented the Levy-Ito model (LIM), directly extending the SDE-based framework to a class of heavy-tailed SDEs, where the injected noise followed an $\alpha$-stable distribution -- a rich class of heavy-tailed distributions. Despite its theoretical elegance and performance improvements, LIM relies on highly involved mathematical techniques, which may limit its accessibility and hinder its broader adoption and further development. In this study, we take a step back, and instead of starting from the SDE formulation, we extend the denoising diffusion probabilistic model (DDPM) by directly replacing the Gaussian noise with $\alpha$-stable noise. By using only elementary proof techniques, we show that the proposed approach, \emph{denoising L\'{e}vy probabilistic model} (DLPM) algorithmically boils down to running vanilla DDPM with minor modifications, hence allowing the use of existing implementations with minimal changes. Remarkably, as opposed to the Gaussian case, DLPM and LIM yield different training algorithms and different backward processes, leading to distinct sampling algorithms. This fundamental difference translates favorably for the performance of DLPM in various aspects: our experiments show that DLPM achieves better coverage of the tails of the data distribution, better generation of unbalanced datasets, and improved computation times requiring significantly smaller number of backward steps.

This paper investigates test-time adaptation (TTA) for regression, where a regression model pre-trained in a source domain is adapted to an unknown target distribution with unlabeled target data. Although regression is one of the fundamental tasks in machine learning, most of the existing TTA methods have classification-specific designs, which assume that models output class-categorical predictions, whereas regression models typically output only single scalar values. To enable TTA for regression, we adopt a feature alignment approach, which aligns the feature distributions between the source and target domains to mitigate the domain gap. However, we found that naive feature alignment employed in existing TTA methods for classification is ineffective or even worse for regression because the features are distributed in a small subspace and many of the raw feature dimensions have little significance to the output. For an effective feature alignment in TTA for regression, we propose Significant-subspace Alignment (SSA). SSA consists of two components: subspace detection and dimension weighting. Subspace detection finds the feature subspace that is representative and significant to the output. Then, the feature alignment is performed in the subspace during TTA. Meanwhile, dimension weighting raises the importance of the dimensions of the feature subspace that have greater significance to the output. We experimentally show that SSA outperforms various baselines on real-world datasets. The code is available at https://github.com/kzkadc/regression-tta.

The heavy-ball momentum method accelerates gradient descent with a momentum term but lacks accelerated convergence for general smooth strongly convex problems. This work introduces the Accelerated Over-Relaxation Heavy-Ball (AOR-HB) method, the first variant with provable global and accelerated convergence for such problems. AOR-HB closes a long-standing theoretical gap, extends to composite convex optimization and min-max problems, and achieves optimal complexity bounds. It offers three key advantages: (1) broad generalization ability, (2) potential to reshape acceleration techniques, and (3) conceptual clarity and elegance compared to existing methods.

We present locality-sensitive avatar, a neural radiance field (NeRF) based network to learn human motions from monocular videos. To this end, we estimate a canonical representation between different frames of a video with a non-linear mapping from observation to canonical space, which we decompose into a skeletal rigid motion and a non-rigid counterpart. Our key contribution is to retain fine-grained details by modeling the non-rigid part with a graph neural network (GNN) that keeps the pose information local to neighboring body parts. Compared to former canonical representation based methods which solely operate on the coordinate space of a whole shape, our locality-sensitive motion modeling can reproduce both realistic shape contours and vivid fine-grained details. We evaluate on ZJU-MoCap, SynWild, ActorsHQ, MVHumanNet and various outdoor videos. The experiments reveal that with the locality sensitive deformation to canonical feature space, we are the first to achieve state-of-the-art results across novel view synthesis, novel pose animation and 3D shape reconstruction simultaneously. Our code is available at https://github.com/ChunjinSong/lsavatar.

In this paper, we introduce Knowledge-Orthogonal Reasoning (KOR), a concept aimed at minimizing reliance on domain-specific knowledge, enabling more accurate evaluation of models' reasoning abilities in out-of-distribution settings. Based on this concept, we propose the Knowledge-Orthogonal Reasoning Benchmark (KOR-Bench), encompassing five task categories: Operation, Logic, Cipher, Puzzle, and Counterfactual. KOR-Bench emphasizes models' effectiveness in applying new rule descriptions to solve novel rule-driven questions. O1-Preview and O1-Mini achieve accuracies of 72.88\% and 70.16\%, surpassing Claude-3.5-Sonnet and GPT-4o (58.96\% and 58.00\%), highlighting the effectiveness of KOR-Bench. We perform detailed analyses, identifying bottlenecks in the Cipher task with Stepwise Prompting, where two rounds of Self-Correction yield optimal results. We evaluate performance across three integrated tasks, explore the impact of Tricks on the Puzzle task, and visualize rule-focused attention. Additionally, we conduct an ablation study on dataset size, benchmark correlations, and zero-shot and three-shot "only questions" experiments. KOR-Bench aims to enhance reasoning evaluation and support further research in this area.

We introduce refined variants of the Local Learning Coefficient (LLC), a measure of model complexity grounded in singular learning theory, to study the development of internal structure in transformer language models during training. By applying these refined LLCs (rLLCs) to individual components of a two-layer attention-only transformer, we gain novel insights into the progressive differentiation and specialization of attention heads. Our methodology reveals how attention heads differentiate into distinct functional roles over the course of training, analyzes the types of data these heads specialize to process, and discovers a previously unidentified multigram circuit. These findings demonstrate that rLLCs provide a principled, quantitative toolkit for developmental interpretability, which aims to understand models through their evolution across the learning process. This work advances the field of developmental interpretability by providing a mathematically rigorous approach to understanding neural networks through the lens of their learning process. More broadly, this work takes a step towards establishing the correspondence between data distributional structure, geometric properties of the loss landscape, learning dynamics, and emergent computational structures in neural networks.

Conservative Contextual Bandits (CCBs) address safety in sequential decision making by requiring that an agent's policy, along with minimizing regret, also satisfies a safety constraint: the performance is not worse than a baseline policy (e.g., the policy that the company has in production) by more than $(1+\alpha)$ factor. Prior work developed UCB-style algorithms for this problem in the multi-armed (Wu et al., 2016) and contextual linear (Kazerouni et al., 2017) settings. However, in practice the cost of the arms is often a non-linear function, and therefore existing UCB algorithms are ineffective in such settings. In this paper, we consider CCBs beyond the linear case and develop two algorithms $\mathtt{C\text{-}SquareCB}$ and $\mathtt{C\text{-}FastCB}$, using Inverse Gap Weighting (IGW) based exploration and an online regression oracle. We show that the safety constraint is satisfied in high probability and that the regret for $\mathtt{C\text{-}SquareCB}$ is sub-linear in horizon $T$, while the the regret for $\mathtt{C\text{-}FastCB}$ is first-order and is sub-linear in $L^*$, the cumulative loss of the optimal policy. Subsequently, we use a neural network for function approximation and online gradient descent as the regression oracle to provide $\tilde{\mathcal{O}}\big(\sqrt{KT} + K/\alpha\big) $ and $\tilde{\mathcal{O}}\big(\sqrt{KL^*} + K (1 + 1/\alpha)\big)$ regret bounds respectively. Finally, we demonstrate the efficacy of our algorithms on real world data, and show that they significantly outperform the existing baseline while maintaining the performance guarantee.

In-context learning (ICL) allows LLMs to learn from examples without changing their weights: this is a particularly promising capability for long-context LLMs that can potentially learn from many examples. Recently, Lin et al. (2024) proposed URIAL, a method using only three in-context examples to align base LLMs, achieving non-trivial instruction following performance. In this work, we show that, while effective, ICL alignment with URIAL still underperforms compared to instruction fine-tuning on established benchmarks such as MT-Bench and AlpacaEval 2.0 (LC), especially with more capable base LLMs. We then uncover the most relevant elements for successful in-context alignment, finding the crucial role of the decoding parameters. Based on these insights, we show that the approach of URIAL can indeed be improved by adding more, potentially carefully selected, high-quality demonstrations in context, getting closer to the performance of instruct models. Finally, we provide the first, to our knowledge, systematic comparison of ICL and instruction fine-tuning (IFT) for instruction following in the low data regime. Overall, our work advances the understanding of ICL as an alignment technique and its relationship to IFT.