ICML 2025 Papers

Likelihood approximations for images are not trivial to compute and can be useful in many applications. We examine the use of Contrastive Language-Image Pre-training (CLIP) to assess the likelihood of images and captions. We introduce Whitened CLIP, a novel transformation of the CLIP latent space via an invertible linear operation. This transformation ensures that each feature in the embedding space has zero mean, unit standard deviation, and no correlation with all other features, resulting in an identity covariance matrix. We show that the whitened embedding statistics can be well approximated by a standard normal distribution, allowing log-likelihood to be estimated using the squared Euclidean norm in the whitened space. The whitening procedure is completely training-free and uses a precomputed whitening matrix, making it extremely fast. We present several preliminary experiments demonstrating the properties and applicability of these likelihood scores to images and captions. Our code is available at github.com/rbetser/W_CLIP/tree/main.

Wearable devices record physiological and behavioral signals that can improve health predictions. While foundation models are increasingly used for such predictions, they have been primarily applied to low-level sensor data, despite behavioral data often being more informative due to their alignment with physiologically relevant timescales and quantities. We develop foundation models of such behavioral signals using over 2.5B hours of wearable data from 162K individuals, systematically optimizing architectures and tokenization strategies for this unique dataset. Evaluated on 57 health-related tasks, our model shows strong performance across diverse real-world applications including individual-level classification and time-varying health state prediction. The model excels in behavior-driven tasks like sleep prediction, and improves further when combined with representations of raw sensor data. These results underscore the importance of tailoring foundation model design to wearables and demonstrate the potential to enable new health applications.

We introduce *Lavender*, a simple supervised fine-tuning (SFT) method that boosts the performance of advanced vision-language models (VLMs) by leveraging state-of-the-art image generation models such as Stable Diffusion. Specifically, Lavender aligns the text-vision attention in the VLM transformer with the equivalent used by Stable Diffusion during SFT, instead of adapting separate encoders. This alignment enriches the model’s visual understanding and significantly boosts performance across in- and out-of-distribution tasks. Lavender requires just 0.13 million training examples---2.5\% of typical large-scale SFT datasets---and fine-tunes on standard hardware (8 GPUs) in a single day. It consistently improves state-of-the-art open-source multimodal LLMs (e.g., Llama-3.2-11B, MiniCPM-Llama3-v2.5), achieving up to 30\% gains and a 68\% boost on challenging out-of-distribution medical QA tasks. By efficiently transferring the visual expertise of image generators with minimal supervision, Lavender offers a scalable solution for more accurate vision-language systems. Code, training data, and models are available on the [project page](https://astrazeneca.github.io/vlm/).

Scaling video generation models is believed to be promising in building world models that adhere to fundamental physical laws. However, whether these models can discover physical laws purely from vision can be questioned. A world model learning the true law should give predictions robust to nuances and correctly extrapolate on unseen scenarios. In this work, we evaluate across three key scenarios: in-distribution, out-of-distribution, and combinatorial generalization. We developed a 2D simulation testbed for object movement and collisions to generate videos deterministically governed by one or more classical mechanics laws. We focus on the scaling behavior of training diffusion-based video generation models to predict object movements based on initial frames. Our scaling experiments show perfect generalization within the distribution, measurable scaling behavior for combinatorial generalization, but failure in out-of-distribution scenarios. Further experiments reveal two key insights about the generalization mechanisms of these models: (1) the models fail to abstract general physical rules and instead exhibit "case-based" generalization behavior, i.e., mimicking the closest training example; (2) when generalizing to new cases, models are observed to prioritize different factors when referencing training data: color $>$ size $>$ velocity $>$ shape. Our study suggests that scaling alone is insufficient for video generation models to uncover fundamental physical laws.

Foundation models are pretrained on large-scale corpora to learn generalizable patterns across domains and tasks---such as contours, textures, and edges in images, or tokens and sentences in text. In contrast, discovering such generalities in graph-structured data, especially across heterogeneous graph tasks, remains an open challenge. To address this, we propose a novel approach to cross-task generalization in graphs via task-trees, which serve as unified learning instances aligning node-, edge-, and graph-level tasks. We theoretically analyze the stability, transferability, and generalization properties of task-trees, showing that pretraining a graph neural network (GNN) on diverse task-trees with a reconstruction objective induces transferable knowledge. This enables efficient adaptation to downstream tasks with minimal fine-tuning. To validate our framework, we introduce Graph Generality Identifier on Task-Trees (GIT), a graph foundation model that demonstrates strong performance on over 30 graphs across five domains via fine-tuning, in-context learning, and zero-shot generalization. Code and data are available at https://github.com/Zehong-Wang/GIT.

Diffusion models have emerged as mainstream framework in visual generation. Building upon this success, the integration of Mixture of Experts (MoE) methods has shown promise in enhancing model scalability and performance. In this paper, we introduce Race-DiT, a novel MoE model for diffusion transformers with a flexible routing strategy, Expert Race. By allowing tokens and experts to compete together and select the top candidates, the model learns to dynamically assign experts to critical tokens. Additionally, we propose per-layer regularization to address challenges in shallow layer learning, and router similarity loss to prevent mode collapse, ensuring better expert utilization. Extensive experiments on ImageNet validate the effectiveness of our approach, showcasing significant performance gains while promising scaling properties.

Large Vision-Language Models (LVLMs) can reason effectively over both textual and visual inputs, but they tend to hallucinate syntactically coherent yet visually ungrounded contents. In this paper, we investigate the internal dynamics of hallucination by examining the tokens logits rankings throughout the generation process, revealing three key patterns in how LVLMs process information: (1) *gradual visual information loss* -- visually grounded tokens gradually become less favored throughout generation, and (2) *early excitation* -- semantically meaningful tokens achieve peak activation in the layers earlier than the final layer. (3) *hidden genuine information* -- visually grounded tokens though not being eventually decided still retain relatively high rankings at inference. Based on these insights, we propose **VISTA** (**V**isual **I**nformation **S**teering with **T**oken-logit **A**ugmentation), a training-free inference-time intervention framework that reduces hallucination while promoting genuine information. VISTA works by combining two complementary approaches: reinforcing visual information in activation space and leveraging early layer activations to promote semantically meaningful decoding. Compared to existing methods, VISTA requires no external supervision and is applicable to various decoding strategies. Extensive experiments show that VISTA on average reduces hallucination by about 40% on evaluated open-ended generation task, and it consistently outperforms existing methods on four benchmarks across four architectures under three decoding strategies. Code is available at https://github.com/LzVv123456/VISTA.

Pre-trained 3D vision models have gained significant attention for their promising performance on point cloud data. However, fully fine-tuning these models for downstream tasks is computationally expensive and storage-intensive. Existing parameter-efficient fine-tuning (PEFT) approaches, which focus primarily on input token prompting, struggle to achieve competitive performance due to their limited ability to capture the geometric information inherent in point clouds. To address this challenge, we propose a novel Geometry-Aware Point Cloud Prompt (GAPrompt) that leverages geometric cues to enhance the adaptability of 3D vision models. First, we introduce a Point Prompt that serves as an auxiliary input alongside the original point cloud, explicitly guiding the model to capture fine-grained geometric details. Additionally, we present a Point Shift Prompter designed to extract global shape information from the point cloud, enabling instance-specific geometric adjustments at the input level. Moreover, our proposed Prompt Propagation mechanism incorporates the shape information into the model's feature extraction process, further strengthening its ability to capture essential geometric characteristics. Extensive experiments demonstrate that GAPrompt significantly outperforms state-of-the-art PEFT methods and achieves competitive results compared to full fine-tuning on various benchmarks, while utilizing only 2.19\% of trainable parameters.

We introduce *GraphGPT*, a novel self-supervised *generative pre-trained* model for graph learning based on the *Graph Eulerian Transformer* (**GET**). First, we propose **GET**, which combines a standard transformer encoder or decoder architecture with an innovative graph-to-sequence transformation method. This method converts graphs or sampled subgraphs into sequences of tokens representing nodes, edges, and attributes in a reversible manner using Eulerian paths. We pre-train **GET** using either of the two self-supervised tasks: next-token prediction (NTP) and scheduled masked-token prediction (SMTP). The pre-trained model is then fine-tuned for downstream tasks such as graph-, edge-, and node-level prediction. Despite its simplicity, GraphGPT achieves performance comparable to or surpassing state-of-the-art methods on multiple large-scale Open Graph Benchmark (OGB) datasets. It demonstrates exceptional results on the molecular property prediction dataset PCQM4Mv2 and the protein-protein interaction dataset ogbl-ppa. Notably, generative pre-training enables scaling GraphGPT to 2 billion parameters while maintaining performance gains — a breakthrough that overcomes the scalability limitations of traditional Graph Neural Networks (GNNs) and prior graph transformers (GTs). To advance research in graph foundation models and facilitate scientific discovery in chemistry, materials science, and related fields, we have released the source code (https://github.com/alibaba/graph-gpt) and model checkpoints (https://www.modelscope.cn/organization/Alibaba-DT).

Despite the empirical success of foundation models, we do not have a systematic characterization of the representations that these models learn. In this paper, we establish the contexture theory. It shows that a large class of representation learning methods can be characterized as learning from the association between the input and a context variable. Specifically, we show that many popular methods aim to approximate the top-d singular functions of the expectation operator induced by the context, in which case we say that the representation learns the contexture. We demonstrate the generality of the contexture theory by proving that representation learning within various learning paradigms -- supervised, self-supervised, and manifold learning -- can all be studied from such a perspective. We prove that representations that learn the contexture are optimal on those tasks that are compatible with the context. One important implication of our theory is that once the model is large enough to approximate the top singular functions, scaling up the model size yields diminishing returns, so further improvement requires better contexts. To this end, we study how to evaluate a context without knowing the downstream tasks. We propose a metric and show by experiments that it correlates well with the actual performance of the encoder on many real datasets.

Long-form video processing fundamentally challenges vision-language models (VLMs) due to the high computational costs of handling extended temporal sequences. Existing token pruning and feature merging methods often sacrifice critical temporal dependencies or dilute semantic information. We introduce differential distillation, a principled approach that systematically preserves task-relevant information while suppressing redundancy. Based on this principle, we develop ViLAMP, a hierarchical video-language model that processes hour-long videos at "mixed precision" through two key mechanisms: (1) differential keyframe selection that maximizes query relevance while maintaining temporal distinctiveness at the frame level and (2) differential feature merging that preserves query-salient features in non-keyframes at the patch level. Hence, ViLAMP retains full information in keyframes while reducing non-keyframes to their most salient features, resembling mixed-precision training. Extensive experiments demonstrate ViLAMP's superior performance across five video understanding benchmarks, particularly on long-form content. Notably, ViLAMP can process ultra-long videos (up to 10K frames) on a single NVIDIA A100 GPU, achieving substantial computational efficiency while maintaining state-of-the-art performance. Code and model are available at https://github.com/steven-ccq/ViLAMP.

The long-standing dominance of gradient-boosted decision trees on tabular data is currently challenged by tabular foundation models using In-Context Learning (ICL): setting the training data as context for the test data and predicting in a single forward pass without parameter updates. While TabPFNv2 foundation model excels on tables with up to 10K samples, its alternating column- and row-wise attentions make handling large training sets computationally prohibitive. So, can ICL be effectively scaled and deliver a benefit for larger tables? We introduce TabICL, a tabular foundation model for classification, pretrained on synthetic datasets with up to 60K samples and capable of handling 500K samples on affordable resources. This is enabled by a novel two-stage architecture: a column-then-row attention mechanism to build fixed-dimensional embeddings of rows, followed by a transformer for efficient ICL. Across 200 classification datasets from the TALENT benchmark, TabICL is on par with TabPFNv2 while being systematically faster (up to 10 times), and significantly outperforms all other approaches. On 53 datasets with over 10K samples, TabICL surpasses both TabPFNv2 and CatBoost, demonstrating the potential of ICL for large data. Pretraining code, inference code, and pre-trained models are available at https://github.com/soda-inria/tabicl.