The diversity of recommendation is equally crucial as accuracy in improving user experience. Existing studies, e.g., Determinantal Point Process (DPP) and Maximal Marginal Relevance (MMR), employ a greedy paradigm to iteratively select items that optimize both accuracy and diversity. However, prior methods typically exhibit quadratic complexity, limiting their applications to the re-ranking stage and are not applicable to other recommendation stages with a larger pool of candidate items, such as the pre-ranking and ranking stages. In this paper, we propose Contextual Distillation Model (CDM), an efficient recommendation model that addresses diversification, suitable for the deployment in all stages of industrial recommendation pipelines. Specifically, CDM utilizes the candidate items in the same user request as context to enhance the diversification of the results. We propose a contrastive context encoder that employs attention mechanisms to model both positive and negative contexts. For the training of CDM, we compare each target item with its context embedding and utilize the knowledge distillation framework to learn the win probability of each target item under the MMR algorithm, where the teacher is derived from MMR outputs. During inference, ranking is performed through a linear combination of the recommendation and student model scores, ensuring both diversity and efficiency. We perform offline evaluations on two industrial datasets and conduct online A/B test of CDM on the short-video platform KuaiShou. The considerable enhancements observed in both recommendation quality and diversity, as shown by metrics, provide strong superiority for the effectiveness of CDM.
In this paper, we study cascading failures in power grids through the lens of information diffusion models. Similar to the spread of rumors or influence in an online social network, it has been observed that failures (outages) in a power grid can spread contagiously, driven by viral spread mechanisms. We employ a stochastic diffusion model that is Markovian (memoryless) and local (the activation of one node, i.e., transmission line, can only be caused by its neighbors). Our model integrates viral diffusion principles with physics-based concepts, by correlating the diffusion weights (contagion probabilities between transmission lines) with the hyperparametric Information Cascades (IC) model. We show that this diffusion model can be learned from traces of cascading failures, enabling accurate modeling and prediction of failure propagation. This approach facilitates actionable information through well-understood and efficient graph analysis methods and graph diffusion simulations. Furthermore, by leveraging the hyperparametric model, we can predict diffusion and mitigate the risks of cascading failures even in unseen grid configurations, whereas existing methods falter due to a lack of training data. Extensive experiments based on a benchmark power grid and simulations therein show that our approach effectively captures the failure diffusion phenomena and guides decisions to strengthen the grid, reducing the risk of large-scale cascading failures. Additionally, we characterize our model's sample complexity, improving upon the existing bound.
This paper proposes a theory of stock market predictability patterns based on a model of heterogeneous beliefs. In a discrete finite time framework, some agents receive news about an asset's fundamental value through a noisy signal. The investors are heterogeneous in that they have different beliefs about the stochastic supply. A momentum in the stock price arises from those agents who incorrectly underestimate the signal accuracy, dampening the initial price impact of the signal. A reversal in price occurs because the price reverts to the fundamental value in the long run. An extension of the model to multiple assets case predicts co-movement and lead-lag effect, in addition to cross-sectional momentum and reversal. The heterogeneous beliefs of investors about news demonstrate how the main predictability anomalies arise endogenously in a model of bounded rationality.
Large model training often uses recomputation to alleviate memory pressure and pipelines to exploit the parallelism of data, tensors, and devices. However, existing recomputation approaches may incur high overhead when training real-world models, as they are executed on demand in the critical training path. In this paper, we present Lynx, a new recomputation framework to reduce overhead by overlapping recomputation with communication in training pipelines. To reduce the large search space for recomputation strategies, we propose a heuristic-based recomputation scheduling algorithm, which is based on the observation that there are identical structures in large DNN models so that we can apply the same scheduling policy to all such structures. Additionally, we propose a recomputation-aware model partitioning method to balance each stage's execution time for improved training throughput. Our comprehensive evaluation using GPT models with 1.3B-23B parameters shows that Lynx outperforms existing recomputation approaches by up to 1.37x.
This paper evaluates the generalization ability of classification models on out-of-distribution test sets without depending on ground truth labels. Common approaches often calculate an unsupervised metric related to a specific model property, like confidence or invariance, which correlates with out-of-distribution accuracy. However, these metrics are typically computed for each test sample individually, leading to potential issues caused by spurious model responses, such as overly high or low confidence. To tackle this challenge, we propose incorporating responses from neighboring test samples into the correctness assessment of each individual sample. In essence, if a model consistently demonstrates high correctness scores for nearby samples, it increases the likelihood of correctly predicting the target sample, and vice versa. The resulting scores are then averaged across all test samples to provide a holistic indication of model accuracy. Developed under the vicinal risk formulation, this approach, named vicinal risk proxy (VRP), computes accuracy without relying on labels. We show that applying the VRP method to existing generalization indicators, such as average confidence and effective invariance, consistently improves over these baselines both methodologically and experimentally. This yields a stronger correlation with model accuracy, especially on challenging out-of-distribution test sets.
Large policies pretrained on a combination of Internet-scale vision-language data and diverse robot demonstrations have the potential to change how we teach robots new skills: rather than training new behaviors from scratch, we can fine-tune such vision-language-action (VLA) models to obtain robust, generalizable policies for visuomotor control. Yet, widespread adoption of VLAs for robotics has been challenging as 1) existing VLAs are largely closed and inaccessible to the public, and 2) prior work fails to explore methods for efficiently fine-tuning VLAs for new tasks, a key component for adoption. Addressing these challenges, we introduce OpenVLA, a 7B-parameter open-source VLA trained on a diverse collection of 970k real-world robot demonstrations. OpenVLA builds on a Llama 2 language model combined with a visual encoder that fuses pretrained features from DINOv2 and SigLIP. As a product of the added data diversity and new model components, OpenVLA demonstrates strong results for generalist manipulation, outperforming closed models such as RT-2-X (55B) by 16.5% in absolute task success rate across 29 tasks and multiple robot embodiments, with 7x fewer parameters. We further show that we can effectively fine-tune OpenVLA for new settings, with especially strong generalization results in multi-task environments involving multiple objects and strong language grounding abilities, and outperform expressive from-scratch imitation learning methods such as Diffusion Policy by 20.4%. We also explore compute efficiency; as a separate contribution, we show that OpenVLA can be fine-tuned on consumer GPUs via modern low-rank adaptation methods and served efficiently via quantization without a hit to downstream success rate. Finally, we release model checkpoints, fine-tuning notebooks, and our PyTorch codebase with built-in support for training VLAs at scale on Open X-Embodiment datasets.
The denoising model has been proven a powerful generative model but has little exploration of discriminative tasks. Representation learning is important in discriminative tasks, which is defined as "learning representations (or features) of the data that make it easier to extract useful information when building classifiers or other predictors". In this paper, we propose a novel Denoising Model for Representation Learning (DenoiseRep) to improve feature discrimination with joint feature extraction and denoising. DenoiseRep views each embedding layer in a backbone as a denoising layer, processing the cascaded embedding layers as if we are recursively denoise features step-by-step. This unifies the frameworks of feature extraction and denoising, where the former progressively embeds features from low-level to high-level, and the latter recursively denoises features step-by-step. After that, DenoiseRep fuses the parameters of feature extraction and denoising layers, and theoretically demonstrates its equivalence before and after the fusion, thus making feature denoising computation-free. DenoiseRep is a label-free algorithm that incrementally improves features but also complementary to the label if available. Experimental results on various discriminative vision tasks, including re-identification (Market-1501, DukeMTMC-reID, MSMT17, CUHK-03, vehicleID), image classification (ImageNet, UB200, Oxford-Pet, Flowers), object detection (COCO), image segmentation (ADE20K) show stability and impressive improvements. We also validate its effectiveness on the CNN (ResNet) and Transformer (ViT, Swin, Vmamda) architectures.
Vision-Language Foundation Models (VLFMs) have made remarkable progress on various multimodal tasks, such as image captioning, image-text retrieval, visual question answering, and visual grounding. However, most methods rely on training with general image datasets, and the lack of geospatial data leads to poor performance on earth observation. Numerous geospatial image-text pair datasets and VLFMs fine-tuned on them have been proposed recently. These new approaches aim to leverage large-scale, multimodal geospatial data to build versatile intelligent models with diverse geo-perceptive capabilities, which we refer to as Vision-Language Geo-Foundation Models (VLGFMs). This paper thoroughly reviews VLGFMs, summarizing and analyzing recent developments in the field. In particular, we introduce the background and motivation behind the rise of VLGFMs, highlighting their unique research significance. Then, we systematically summarize the core technologies employed in VLGFMs, including data construction, model architectures, and applications of various multimodal geospatial tasks. Finally, we conclude with insights, issues, and discussions regarding future research directions. To the best of our knowledge, this is the first comprehensive literature review of VLGFMs. We keep tracing related works at https://github.com/zytx121/Awesome-VLGFM.
Foundation Models (FMs) serve as a general class for the development of artificial intelligence systems, offering broad potential for generalization across a spectrum of downstream tasks. Despite extensive research into self-supervised learning as the cornerstone of FMs, several outstanding issues persist in Graph Foundation Models that rely on graph self-supervised learning, namely: 1) Homogenization. The extent of generalization capability on downstream tasks remains unclear. 2) Scalability. It is unknown how effectively these models can scale to large datasets. 3) Efficiency. The training time and memory usage of these models require evaluation. 4) Training Stop Criteria. Determining the optimal stopping strategy for pre-training across multiple tasks to maximize performance on downstream tasks. To address these questions, we have constructed a rigorous benchmark that thoroughly analyzes and studies the generalization and scalability of self-supervised Graph Neural Network (GNN) models. Regarding generalization, we have implemented and compared the performance of various self-supervised GNN models, trained to generate node representations, across tasks such as node classification, link prediction, and node clustering. For scalability, we have compared the performance of various models after training using full-batch and mini-batch strategies. Additionally, we have assessed the training efficiency of these models by conducting experiments to test their GPU memory usage and throughput. Through these experiments, we aim to provide insights to motivate future research. The code for this benchmark is publicly available at https://github.com/NYUSHCS/GraphFM.
Deep neural network training often involves stochastic optimization, meaning each run will produce a different model. This implies that hyperparameters of the training process, such as the random seed itself, can potentially have significant influence on the variability in the trained models. Measuring model quality by summary statistics, such as test accuracy, can obscure this dependence. We propose a robust hypothesis testing framework and a novel summary statistic, the $\alpha$-trimming level, to measure model similarity. Applying hypothesis testing directly with the $\alpha$-trimming level is challenging because we cannot accurately describe the distribution under the null hypothesis. Our framework addresses this issue by determining how closely an approximate distribution resembles the expected distribution of a group of individually trained models and using this approximation as our reference. We then use the $\alpha$-trimming level to suggest how many training runs should be sampled to ensure that an ensemble is a reliable representative of the true model performance. We also show how to use the $\alpha$-trimming level to measure model variability and demonstrate experimentally that it is more expressive than performance metrics like validation accuracy, churn, or expected calibration error when taken alone. An application of fine-tuning over random seed in transfer learning illustrates the advantage of our new metric.
During modeling of dynamical systems, often two or more model architectures are combined to obtain a more powerful or efficient model regarding a specific application area. This covers the combination of multiple machine learning architectures, as well as hybrid models, i.e., the combination of physical simulation models and machine learning. In this work, we briefly discuss which types of model are usually combined in dynamical systems modeling and propose a class of models that is capable of expressing mixed algebraic, discrete, and differential equation-based models. Further, we examine different established, as well as new ways of combining these models from the point of view of system theory and highlight two challenges - algebraic loops and local event functions in discontinuous models - that require a special approach. Finally, we propose a new wildcard architecture that is capable of describing arbitrary combinations of models in an easy-to-interpret fashion that can be learned as part of a gradient-based optimization procedure. In a final experiment, different combination architectures between two models are learned, interpreted, and compared using the methodology and software implementation provided.
In recent years, considerable work has been devoted to explaining predictive, deep learning-based models, and in turn how to evaluate explanations. An important class of evaluation methods are ones that are human-centered, which typically require the communication of explanations through visualizations. And while visualization plays a critical role in perceiving and understanding model explanations, how visualization design impacts human perception of explanations remains poorly understood. In this work, we study the graphical perception of model explanations, specifically, saliency-based explanations for visual recognition models. We propose an experimental design to investigate how human perception is influenced by visualization design, wherein we study the task of alignment assessment, or whether a saliency map aligns with an object in an image. Our findings show that factors related to visualization design decisions, the type of alignment, and qualities of the saliency map all play important roles in how humans perceive saliency-based visual explanations.
Greybox fuzzing has achieved success in revealing bugs and vulnerabilities in programs. However, randomized mutation strategies have limited the fuzzer's performance on structured data. Specialized fuzzers can handle complex structured data, but require additional efforts in grammar and suffer from low throughput. In this paper, we explore the potential of utilizing the Large Language Model to enhance greybox fuzzing for structured data. We utilize the pre-trained knowledge of LLM about data conversion and format to generate new valid inputs. We further fine-tuned it with paired mutation seeds to learn structured format and mutation strategies effectively. Our LLM-based fuzzer, LLAMAFUZZ, integrates the power of LLM to understand and mutate structured data to fuzzing. We conduct experiments on the standard bug-based benchmark Magma and a wide variety of real-world programs. LLAMAFUZZ outperforms our top competitor by 41 bugs on average. We also identified 47 unique bugs across all trials. Moreover, LLAMAFUZZ demonstrated consistent performance on both bug trigger and bug reached. Compared to AFL++, LLAMAFUZZ achieved 27.19% more branches in real-world program sets on average. We also demonstrate a case study to explain how LLMs enhance the fuzzing process in terms of code coverage.
Progress in AI is often demonstrated by new models claiming improved performance on tasks measuring model capabilities. Evaluating language models in particular is challenging, as small changes to how a model is evaluated on a task can lead to large changes in measured performance. There is no common standard setup, so different models are evaluated on the same tasks in different ways, leading to claims about which models perform best not being reproducible. We propose OLMES, a completely documented, practical, open standard for reproducible LLM evaluations. In developing this standard, we identify and review the varying factors in evaluation practices adopted by the community - such as details of prompt formatting, choice of in-context examples, probability normalizations, and task formulation. In particular, OLMES supports meaningful comparisons between smaller base models that require the unnatural "cloze" formulation of multiple-choice questions against larger models that can utilize the original formulation. OLMES includes well-considered recommendations guided by results from existing literature as well as new experiments investigating open questions.
We formulate SVBRDF estimation from photographs as a diffusion task. To model the distribution of spatially varying materials, we first train a novel unconditional SVBRDF diffusion backbone model on a large set of 312,165 synthetic spatially varying material exemplars. This SVBRDF diffusion backbone model, named MatFusion, can then serve as a basis for refining a conditional diffusion model to estimate the material properties from a photograph under controlled or uncontrolled lighting. Our backbone MatFusion model is trained using only a loss on the reflectance properties, and therefore refinement can be paired with more expensive rendering methods without the need for backpropagation during training. Because the conditional SVBRDF diffusion models are generative, we can synthesize multiple SVBRDF estimates from the same input photograph from which the user can select the one that best matches the users' expectation. We demonstrate the flexibility of our method by refining different SVBRDF diffusion models conditioned on different types of incident lighting, and show that for a single photograph under colocated flash lighting our method achieves equal or better accuracy than existing SVBRDF estimation methods.
To improve the predictive capacity of system models in the input-output sense, this paper presents a framework for model updating via learning of modeling uncertainties in locally (and thus also in globally) Lipschitz nonlinear systems. First, we introduce a method to extend an existing known model with an uncertainty model so that stability of the extended model is guaranteed in the sense of set invariance and input-to-state stability. To achieve this, we provide two tractable semi-definite programs. These programs allow obtaining optimal uncertainty model parameters for both locally and globally Lipschitz nonlinear models, given uncertainty and state trajectories. Subsequently, in order to extract this data from the available input-output trajectories, we introduce a filter that incorporates an approximated internal model of the uncertainty and asymptotically estimates uncertainty and state realizations. This filter is also synthesized using semi-definite programs with guaranteed robustness with respect to uncertainty model mismatches, disturbances, and noise. Numerical simulations for a large data-set of a roll plane model of a vehicle illustrate the effectiveness and practicality of the proposed methodology in improving model accuracy, while guaranteeing stability.
Contact between languages has the potential to transmit vocabulary and other language features; however, this does not always happen. Here, an iterated learning model is used to examine, in a simple way, the resistance of languages to change during language contact. Iterated learning models are agent-based models of language change, they demonstrate that languages that are expressive and compositional arise spontaneously as a consequence of a language transmission bottleneck. A recently introduced type of iterated learning model, the Semi-Supervised ILM is used to simulate language contact. These simulations do not include many of the complex factors involved in language contact and do not model a population of speakers; nonetheless the model demonstrates that the dynamics which lead languages in the model to spontaneously become expressive and compositional, also cause a language to maintain its core traits even after mixing with another language.
Real-world applications of machine learning models are often subject to legal or policy-based regulations. Some of these regulations require ensuring the validity of the model, i.e., the approximation error being smaller than a threshold. A global metric is generally too insensitive to determine the validity of a specific prediction, whereas evaluating local validity is costly since it requires gathering additional data.We propose learning the model error to acquire a local validity estimate while reducing the amount of required data through active learning. Using model validation benchmarks, we provide empirical evidence that the proposed method can lead to an error model with sufficient discriminative properties using a relatively small amount of data. Furthermore, an increased sensitivity to local changes of the validity bounds compared to alternative approaches is demonstrated.
Large language models (LLMs) have achieved remarkable performance on Natural Language Processing (NLP) tasks, but they are hindered by high computational costs and memory requirements. Ternarization, an extreme form of quantization, offers a solution by reducing memory usage and enabling energy-efficient floating-point additions. However, applying ternarization to LLMs faces challenges stemming from outliers in both weights and activations. In this work, observing asymmetric outliers and non-zero means in weights, we introduce Dual Learnable Ternarization (DLT), which enables both scales and shifts to be learnable. We also propose Outlier-Friendly Feature Knowledge Distillation (OFF) to recover the information lost in extremely low-bit quantization. The proposed OFF can incorporate semantic information and is insensitive to outliers. At the core of OFF is maximizing the mutual information between features in ternarized and floating-point models using cosine similarity. Extensive experiments demonstrate that our TernaryLLM surpasses previous low-bit quantization methods on the standard text generation and zero-shot benchmarks for different LLM families. Specifically, for one of the most powerful open-source models, LLaMA-3, our approach (W1.58A16) outperforms the previous state-of-the-art method (W2A16) by 5.8 in terms of perplexity on C4 and by 8.2% in terms of average accuracy on zero-shot tasks.
Recent breakthroughs in large language model-driven autonomous agents have revealed that multi-agent collaboration often surpasses each individual through collective reasoning. Inspired by the neural scaling law--increasing neurons enhances performance, this study explores whether the continuous addition of collaborative agents can yield similar benefits. Technically, we utilize directed acyclic graphs to organize agents into a multi-agent collaboration network (MacNet), upon which their interactive reasoning is topologically orchestrated for autonomous task solving. Extensive evaluations reveal that it effectively supports collaboration among over a thousand agents, with irregular topologies outperforming regular ones. We also identify a collaborative scaling law--the overall performance follows a logistic growth pattern as agents scale, with collaborative emergence occurring earlier than traditional neural emergence. We speculate this may be because scaling agents catalyzes their multidimensional considerations during interactive reflection and refinement, thereby producing more comprehensive artifacts. The code is available at https://github.com/OpenBMB/ChatDev/tree/macnet.