Merging mechanisms, as a mature business model in the field of online advertising, refers to the practice where platforms sort and display sponsored ads provided by advertisers alongside organic results to users according to specific rules. However, in real-world industrial scenarios, advertisers are gradually adopting autobidding instead of manual bidding—they only need to provide high-level constraints like target Return-on-Spend (tROS) to the agent, which then bids on their behalf to maximize multi-round value. Existing studies often overlook this actual business form, resulting in suboptimal outcomes. Meanwhile, the coexistence of the same item in both ad and organic result forms within merging mechanisms further increases the complexity of the context. In terms of interests, advertisers aim to maximize conversion value, while platforms seek to increase the revenue while ensuring user experience, thereby enhancing reputation. Nevertheless, existing works often fail to address this multi-stakeholder challenge in the modern auto-bidding era. To address these issues, we introduce IMPACTNet, an end-to-end framework based on automated mechanism design that learns a unified allocation and pricing mechanism. IMPACTNet directly incorporates advertisers' tROS constraints, models complex contextual information using a transformer-based architecture, and introduces a learnable, state-aware de-duplication strategy. By formulating the design as a constrained optimization problem, our framework learns a mechanism that ensures Auto-bidding Incentive Compatibility (AIC), ensuring truthfully reporting tROS a dominant strategy. Extensive experiments on synthetic and large-scale industrial datasets demonstrate that IMPACTNet significantly outperforms established baselines, achieving a better balance of platform objectives, user experience, and advertiser tROS satisfaction.

Spatially resolved sequencing technologies have revolutionized our understanding of biological regulatory processes within tissue microenvironments by simultaneously capturing the states of genomic regions, genes, and proteins alongside the spatial organization of cells. However, inherent heterogeneity across modalities and samples poses substantial challenges for the integrative analysis of spatial omics data, underscoring the urgent need for advanced computational methods. In this study, we propose PRESENT, a contrastive learning-based integrative framework for cross-modality representation of spatial multi-omics data. PRESENT employs omics-specific encoders consisting of graph attention networks and Bayesian neural networks coupled with distribution-aware decoders to model distinct modalities, and an inter-omics alignment module for multi-omics integration. By effectively incorporating spatial dependencies with multi-omics information across diverse species and technologies, PRESENT facilitates the accurate identification of spatial domains and the elucidation of underlying regulatory mechanisms. Furthermore, PRESENT can be extended to multi-sample integration via a two-stage training workflow, which incorporates inter-batch alignment loss, intra-batch preserving loss, batch-adversarial learning, and cyclic graph refinement strategies to eliminate batch effects while retaining biological signals. Extensive experiments on tissue samples across different anatomical regions and developmental stages demonstrate that PRESENT enables the characterization of hierarchical tissue structures from a spatiotemporal perspective.

Visual imitation learning methods demonstrate strong performance, yet they lack generalization when faced with visual input perturbations like variations in lighting and textures. This limitation hampers their practical application in real-world settings. To address this, we propose ***Stem-OB*** that leverages the inversion process of pretrained image diffusion models to suppress low-level visual differences while maintaining high-level scene structures. This image inversion process is akin to transforming the observation into a shared representation, from which other observations also stem. *Stem-OB* offers a simple yet effective plug-and-play solution that stands in contrast to data augmentation approaches. It demonstrates robustness to various unspecified appearance changes without the need for additional training. We provide theoretical insights and empirical results that validate the efficacy of our approach in simulated and real settings. *Stem-OB* shows an exceptionally significant improvement in real-world robotic tasks, where challenging light and appearance changes are present, with an average increase of **22.2%** in success rates compared to the best baseline. Please refer to [this link](https://stem-ob.github.io/) for more videos and details.

We present a novel approach, MAGIC (manipulation analogies for generalizable intelligent contacts), for one-shot learning of manipulation strategies with fast and extensive generalization to novel objects. By leveraging a reference action trajectory, MAGIC effectively identifies similar contact points and sequences of actions on novel objects to replicate a demonstrated strategy, such as using different hooks to retrieve distant objects of different shapes and sizes. Our method is based on a twostage contact-point matching process that combines global shape matching using pretrained neural features with local curvature analysis to ensure precise and physically plausible contact points. We experiment with three tasks including scooping, hanging, and hooking objects. MAGIC demonstrates superior performance over existing methods, achieving significant improvements in runtime speed and generalization to different object categories. Website: https://magic-2024.github.io/.

On e-commerce platforms, sellers typically bid for impressions from ad traffic to promote their products. However, for most sellers, the majority of their sales come from organic traffic. Consequently, the relationship between their ad spending and total sales remains uncertain, resulting in operational inefficiency. To address this issue, e-commerce platforms have recently introduced a novel platform-wide marketing service known as QuanZhanTui, which has reportedly enhanced marketing efficiency for sellers and driven substantial revenue growth for platforms. QuanZhanTui allows sellers to bid for impressions from the platform's entire traffic to boost their total sales without compromising the platform's user experience. In this paper, we investigate the mechanism design problem that arises from QuanZhanTui. The problem is formulated as a multi-objective optimization to balance sellers' welfare and platform's user experience. We first introduce the stock-constrained value maximizer model, which reflects sellers' dual requirements on marketing efficiency and platform-wide ROI. Then, we propose the Liquid Payment Auction (LPA), an auction designed to optimize the balanced objectives while accounting for sellers' requirements in the auto-bidding environment. It employs a simple payment rule based on sellers' liquid welfare, providing a clearer link between their investment and total sales. Under mild assumptions, we theoretically prove desirable properties of LPA, such as optimality and incentive compatibility. Extensive experiments demonstrate LPA's superior performance over conventional auctions in QuanZhanTui.

Offline reinforcement learning (RL) aims to find a near-optimal policy using pre-collected datasets. Given recent advances in Large Language Models (LLMs) and their few-shot learning prowess, this paper introduces $\textbf{La}$nguage Models for $\textbf{Mo}$tion Control ($\textbf{LaMo}$), a general framework based on Decision Transformers to effectively use pre-trained Language Models (LMs) for offline RL. Our framework highlights four crucial components: (1) Initializing Decision Transformers with sequentially pre-trained LMs, (2) employing the LoRA fine-tuning method, in contrast to full-weight fine-tuning, to combine the pre-trained knowledge from LMs and in-domain knowledge effectively, (3) using the non-linear MLP transformation instead of linear projections, to generate embeddings, and (4) integrating an auxiliary language prediction loss during fine-tuning to stabilize the LMs and retain their original abilities on languages. Empirical results indicate $\textbf{LaMo}$ achieves state-of-the-art performance in sparse-reward tasks and closes the gap between value-based offline RL methods and decision transformers in dense-reward tasks. In particular, our method demonstrates superior performance in scenarios with limited data samples.

Sequential decision-making algorithms such as reinforcement learning (RL) in real-world scenarios inevitably face environments with partial observability. This paper scrutinizes the effectiveness of a popular architecture, namely Transformers, in Partially Observable Markov Decision Processes (POMDPs) and reveals its theoretical limitations. We establish that regular languages, which Transformers struggle to model, are reducible to POMDPs. This poses a significant challenge for Transformers in learning POMDP-specific inductive biases, due to their lack of inherent recurrence found in other models like RNNs. This paper casts doubt on the prevalent belief in Transformers as sequence models for RL and proposes to introduce a point-wise recurrent structure. The Deep Linear Recurrent Unit (LRU) emerges as a well-suited alternative for Partially Observable RL, with empirical results highlighting the sub-optimal performance of the Transformer and considerable strength of LRU.

The sound speed profile (SSP) reflects the variation of sound speed from the sea surface to the seabed, and its structural changes will have a major impact on sound propagation. There is a mapping link between the ocean’s surface and underwater parameters because of its nonlinear dynamic properties. Accurately inverse SSPs by sea surface remote sensing data use has become a research hotspot in recent years. In this letter, we propose for the first time to use a combination of single empirical orthogonal function (sEOF) and generalized regression neural network (GRNN) to inverse SSPs in the Luzon Strait using satellite remote sensing data, including sea surface temperature anomaly (SSTA), sea level anomaly (SLA), and eddy kinetic energy (EKE). To satisfy training requirements, necessary quality control of the Argo data is performed in the study area. Despite the limited amount of data, sEOF-GRNN can still achieve better inversion results. When sEOF-GRNN is used instead of linear regression (sEOF-R), the performance is improved by more than 50% and the root mean square error (RMSE) is less than 1.5 m/s. Based on Argo-SSPs and inversed SSPs, the differences in sound propagation are analyzed. With the aid of remote sensing data, this study offers theoretical and technical support for the precise inversion of SSPs.

Human hands possess remarkable dexterity and have long served as a source of inspiration for robotic manipulation. In this work, we propose a human $\textbf{H}$and-$\textbf{In}$formed visual representation learning framework to solve difficult $\textbf{Dex}$terous manipulation tasks ($\textbf{H-InDex}$) with reinforcement learning. Our framework consists of three stages: $\textit{(i)}$ pre-training representations with 3D human hand pose estimation, $\textit{(ii)}$ offline adapting representations with self-supervised keypoint detection, and $\textit{(iii)}$ reinforcement learning with exponential moving average BatchNorm. The last two stages only modify $0.36$% parameters of the pre-trained representation in total, ensuring the knowledge from pre-training is maintained to the full extent. We empirically study $\textbf{12}$ challenging dexterous manipulation tasks and find that $\textbf{H-InDex}$ largely surpasses strong baseline methods and the recent visual foundation models for motor control. Code and videos are available at https://yanjieze.com/H-InDex .

Accurate checkerboard recognition is vital for basic camera calibration in many machine vision tasks. However, existing detectors usually encounter two problems. First, X-corner location precision is easily affected by undesired factors such as noise and distortion. Then, the pattern recovery requires manual input of the dimensions of the real corner matrix, and most recovery schemes can’t deal well with the pattern with occlusion or missing corners. In this paper we propose a novel CNN-based checkerboard detection framework to address these problems. This framework consists of three sub-modules: 1) an X-corner detection network to identify as many real corners as possible. 2) a sub-pixel refinement technique obtained from the geometric analysis of the image response map and grayscale to find precise corner locations. 3) a pattern recovery scheme to find the regular checkerboard layout even with high distortion and partial occlusion. Quantitative experimental results show that the proposed approach has higher accuracy and stronger robustness than state-of-the-art methods to both synthetic images and real-world camera calibration scenarios.