Search

Large Language Models (LLMs) are increasingly employed in complex workflows, where different LLMs and fine-tuned variants collaboratively address complex tasks. However, these systems face significant inefficiencies due to redundant context processing of the shared context. We propose DroidSpeak, a framework that optimizes context sharing between fine-tuned LLMs derived from the same foundational model. DroidSpeak identifies critical layers in the KV cache and selectively recomputes them, enabling effective reuse of intermediate data while maintaining high accuracy. Our approach balances computational efficiency and task fidelity, significantly reducing inference latency and throughput bottlenecks. Experiments on diverse datasets and model pairs demonstrate that DroidSpeak achieves up to 3x higher throughputs and 2.6x faster prefill times with negligible accuracy loss compared to full recomputation.

In the digital world, influencers are pivotal as opinion leaders, shaping the views and choices of their influencees. Modern advertising often follows this trend, where marketers choose appropriate influencers for product endorsements, based on thorough market analysis. Previous studies on influencer selection have typically relied on numerical representations of individual opinions and interactions, a method that simplifies the intricacies of social dynamics. In this work, we first introduce a Time-aware Influencer Simulator (TIS), helping promoters identify and select the right influencers to market their products, based on LLM simulation. To validate our approach, we conduct experiments on the public advertising campaign dataset SAGraph which encompasses social relationships, posts, and user interactions. The results show that our method outperforms traditional numerical feature-based approaches and methods using limited LLM agents. Our research shows that simulating user timelines and content lifecycles over time simplifies scaling, allowing for large-scale agent simulations in social networks. Additionally, LLM-based agents for social recommendations and advertising offer substantial benefits for decision-making in promotional campaigns., Comment: 10 pages, 5 figures

With the growing spread of misinformation online, research has increasingly focused on detecting and tracking fake news. However, an overlooked issue is that fake news does not naturally exist in social networks -- it often originates from distorted facts or deliberate fabrication by malicious actors. Understanding how true news gradually evolves into fake news is critical for early detection and prevention, reducing its spread and impact. Hence, in this paper, we take the first step toward simulating and revealing this evolution, proposing a Fake News evolUtion Simulation framEwork (FUSE) based on large language models (LLMs). Specifically, we employ LLM as agents to represent individuals in a simulated social network. We define four types of agents commonly observed in daily interactions: spreaders, who propagate information; commentators, who provide opinions and interpretations; verifiers, who check the accuracy of information; and bystanders, who passively observe without engaging. For simulated environments, we model various social network structures, such as high-clustering networks and scale-free networks, to mirror real-world network dynamics. Each day, the agents engage in belief exchanges, reflect on their thought processes, and reintroduce the news accordingly. Given the lack of prior work in this area, we developed a FUSE-EVAL evaluation framework to measure the deviation from true news during the fake news evolution process. The results show that FUSE successfully captures the underlying patterns of how true news transforms into fake news and accurately reproduces previously discovered instances of fake news, aligning closely with human evaluations. Moreover, our work provides insights into the fact that combating fake news should not be delayed until it has fully evolved; instead, prevention in advance is key to achieving better outcomes.

The impact of initial connectivity on learning has been extensively studied in the context of backpropagation-based gradient descent, but it remains largely underexplored in biologically plausible learning settings. Focusing on recurrent neural networks (RNNs), we found that the initial weight magnitude significantly influences the learning performance of biologically plausible learning rules in a similar manner to its previously observed effect on training via backpropagation through time (BPTT). By examining the maximum Lyapunov exponent before and after training, we uncovered the greater demands that certain initialization schemes place on training to achieve desired information propagation properties. Consequently, we extended the recently proposed gradient flossing method, which regularizes the Lyapunov exponents, to biologically plausible learning and observed an improvement in learning performance. To our knowledge, we are the first to examine the impact of initialization on biologically plausible learning rules for RNNs and to subsequently propose a biologically plausible remedy. Such an investigation could lead to predictions about the influence of initial connectivity on learning dynamics and performance, as well as guide neuromorphic design., Comment: 9 pages, 5 figures

Designing a universal policy architecture that performs well across diverse robots and task configurations remains a key challenge. In this work, we address this by representing robot actions as sequential data and generating actions through autoregressive sequence modeling. Existing autoregressive architectures generate end-effector waypoints sequentially as word tokens in language modeling, which are limited to low-frequency control tasks. Unlike language, robot actions are heterogeneous and often include continuous values -- such as joint positions, 2D pixel coordinates, and end-effector poses -- which are not easily suited for language-based modeling. Based on this insight, we introduce a straightforward enhancement: we extend causal transformers' single-token prediction to support predicting a variable number of tokens in a single step through our Chunking Causal Transformer (CCT). This enhancement enables robust performance across diverse tasks of various control frequencies, greater efficiency by having fewer autoregression steps, and lead to a hybrid action sequence design by mixing different types of actions and using a different chunk size for each action type. Based on CCT, we propose the Autoregressive Policy (ARP) architecture, which solves manipulation tasks by generating hybrid action sequences. We evaluate ARP across diverse robotic manipulation environments, including Push-T, ALOHA, and RLBench, and show that ARP, as a universal architecture, outperforms the environment-specific state-of-the-art in all tested benchmarks, while being more efficient in computation and parameter sizes. Videos of our real robot demonstrations, all source code and the pretrained models of ARP can be found at http://github.com/mlzxy/arp.

In response to the increasing mental health challenges faced by college students, we sought to understand their perspectives on how AI applications, particularly Large Language Models (LLMs), can be leveraged to enhance their mental well-being. Through pilot interviews with ten diverse students, we explored their opinions on the use of LLMs across five fictional scenarios: General Information Inquiry, Initial Screening, Reshaping Patient-Expert Dynamics, Long-term Care, and Follow-up Care. Our findings revealed that students' acceptance of LLMs varied by scenario, with participants highlighting both potential benefits, such as proactive engagement and personalized follow-up care, and concerns, including limitations in training data and emotional support. These insights inform how AI technology should be designed and implemented to effectively support and enhance students' mental well-being, particularly in scenarios where LLMs can complement traditional methods, while maintaining empathy and respecting individual preferences., Comment: 5 pages

This paper considers the problem of testing and estimation of change point where signals after the change point can be highly irregular, which departs from the existing literature that assumes signals after the change point to be piece-wise constant or vary smoothly. A two-step approach is proposed to effectively estimate the location of the change point. The first step consists of a preliminary estimation of the change point that allows us to obtain unknown parameters for the second step. In the second step we use a new procedure to determine the position of the change point. We show that, under suitable conditions, the desirable $\mathcal{O}_P(1)$ rate of convergence of the estimated change point can be obtained. We apply our method to analyze the Baidu search index of COVID-19 related symptoms and find 8~December 2019 to be the starting date of the COVID-19 pandemic., Comment: 18 pages, 6 figures

We propose an optimization technique for 3-D underwater object modeling from 2-D forward-scan sonar images at known poses. A key contribution, for objects imaged in the proximity of the sea surface, is to resolve the multipath artifacts due to the air-water interface. Here, the object image formed by the direct target backscatter is almost always corrupted by the ghost and sometimes by the mirror components (generated by the multipath propagation). Assuming a planar air-water interface, we model, localize, and discard the corrupted object region within each view, thus avoiding the distortion of recovered 3-D shape. Additionally, complementary visual cues from the boundary of the mirror component, distinct at suitable sonar poses, are employed to enhance the 3-D modeling accuracy. The optimization is implemented as iterative shape adjustment by displacing the vertices of triangular patches in the 3-D surface mesh model, in order to minimize the discrepancy between the data and synthesized views of the 3-D object model. To this end, we first determine 2-D motion fields that align the object regions in the data and synthesized views, then calculate the 3-D motion of triangular patch centers, and finally the model vertices. The 3-D model is initialized with the solution of an earlier space carving method applied to the same data. The same parameters are applied in various experiments with 2 real data sets, mixed real-synthetic data set, and computer-generated data guided by general findings from a real experiment, to explore the impact of non-flat air-water interface. The results confirm the generation of a refined 3-D model in about half-dozen iterations., Comment: Copyright 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works

Solving storage problem: where objects must be accurately placed into containers with precise orientations and positions, presents a distinct challenge that extends beyond traditional rearrangement tasks. These challenges are primarily due to the need for fine-grained 6D manipulation and the inherent multi-modality of solution spaces, where multiple viable goal configurations exist for the same storage container. We present a novel Diffusion-based Affordance Prediction (DAP) pipeline for the multi-modal object storage problem. DAP leverages a two-step approach, initially identifying a placeable region on the container and then precisely computing the relative pose between the object and that region. Existing methods either struggle with multi-modality issues or computation-intensive training. Our experiments demonstrate DAP's superior performance and training efficiency over the current state-of-the-art RPDiff, achieving remarkable results on the RPDiff benchmark. Additionally, our experiments showcase DAP's data efficiency in real-world applications, an advancement over existing simulation-driven approaches. Our contribution fills a gap in robotic manipulation research by offering a solution that is both computationally efficient and capable of handling real-world variability. Code and supplementary material can be found at: https://github.com/changhaonan/DPS.git., Comment: Paper Accepted by IROS2024. Arxiv version is 8 pages

We study quantum state testing where the goal is to test whether $\rho=\rho_0\in\mathbb{C}^{d\times d}$ or $\|\rho-\rho_0\|_1>\varepsilon$, given $n$ copies of $\rho$ and a known state description $\rho_0$. In practice, not all measurements can be easily applied, even using unentangled measurements where each copy is measured separately. We develop an information-theoretic framework that yields unified copy complexity lower bounds for restricted families of non-adaptive measurements through a novel measurement information channel. Using this framework, we obtain the optimal bounds for a natural family of $k$-outcome measurements with fixed and randomized schemes. We demonstrate a separation between these two schemes, showing the power of randomized measurement schemes over fixed ones. Previously, little was known for fixed schemes, and tight bounds were only known for randomized schemes with $k\ge d$ and Pauli observables, a special class of 2-outcome measurements. Our work bridges this gap in the literature., Comment: 43 pages. Part of the work was published at COLT 2024. arXiv admin note: text overlap with arXiv:2401.09650

The Sparse Vector Technique (SVT) is one of the most fundamental tools in differential privacy (DP). It works as a backbone for adaptive data analysis by answering a sequence of queries on a given dataset, and gleaning useful information in a privacy-preserving manner. Unlike the typical private query releases that directly publicize the noisy query results, SVT is less informative -- it keeps the noisy query results to itself and only reveals a binary bit for each query, indicating whether the query result surpasses a predefined threshold. To provide a rigorous DP guarantee for SVT, prior works in the literature adopt a conservative privacy analysis by assuming the direct disclosure of noisy query results as in typical private query releases. This approach, however, hinders SVT from achieving higher query accuracy due to an overestimation of the privacy risks, which further leads to an excessive noise injection using the Laplacian or Gaussian noise for perturbation. Motivated by this, we provide a new privacy analysis for SVT by considering its less informative nature. Our analysis results not only broaden the range of applicable noise types for perturbation in SVT, but also identify the exponential noise as optimal among all evaluated noises (which, however, is usually deemed non-applicable in prior works). The main challenge in applying exponential noise to SVT is mitigating the sub-optimal performance due to the bias introduced by noise distributions. To address this, we develop a utility-oriented optimal threshold correction method and an appending strategy, which enhances the performance of SVT by increasing the precision and recall, respectively. The effectiveness of our proposed methods is substantiated both theoretically and empirically, demonstrating significant improvements up to $50\%$ across evaluated metrics.

This study examines fairness within the rideshare industry, focusing on both drivers' wages and riders' trip fares. Through quantitative analysis, we found that drivers' hourly wages are significantly influenced by factors such as race/ethnicity, health insurance status, tenure to the platform, and working hours. Despite platforms' policies not intentionally embedding biases, disparities persist based on these characteristics. For ride fares, we propose a method to audit the pricing policy of a proprietary algorithm by replicating it; we conduct a hypothesis test to determine if the predicted rideshare fare is greater than the taxi fare, taking into account the approximation error in the replicated model. Challenges in accessing data and transparency hinder our ability to isolate discrimination from other factors, underscoring the need for collaboration with rideshare platforms and drivers to enhance fairness in algorithmic wage determination and pricing., Comment: Accepted to the Humans, Algorithmic Decision-Making and Society: Modeling Interactions and Impact, co-located with the International Conference on Machine Learning, Vienna, Austria

The shuffle model of Differential Privacy (DP) is an enhanced privacy protocol which introduces an intermediate trusted server between local users and a central data curator. It significantly amplifies the central DP guarantee by anonymizing and shuffling the local randomized data. Yet, deriving a tight privacy bound is challenging due to its complicated randomization protocol. While most existing work are focused on unified local privacy settings, this work focuses on deriving the central privacy bound for a more practical setting where personalized local privacy is required by each user. To bound the privacy after shuffling, we first need to capture the probability of each user generating clones of the neighboring data points. Second, we need to quantify the indistinguishability between two distributions of the number of clones on neighboring datasets. Existing works either inaccurately capture the probability, or underestimate the indistinguishability between neighboring datasets. Motivated by this, we develop a more precise analysis, which yields a general and tighter bound for arbitrary DP mechanisms. Firstly, we derive the clone-generating probability by hypothesis testing %from a randomizer-specific perspective, which leads to a more accurate characterization of the probability. Secondly, we analyze the indistinguishability in the context of $f$-DP, where the convexity of the distributions is leveraged to achieve a tighter privacy bound. Theoretical and numerical results demonstrate that our bound remarkably outperforms the existing results in the literature.

Homography estimation is the task of determining the transformation from an image pair. Our approach focuses on employing detector-free feature matching methods to address this issue. Previous work has underscored the importance of incorporating semantic information, however there still lacks an efficient way to utilize semantic information. Previous methods suffer from treating the semantics as a pre-processing, causing the utilization of semantics overly coarse-grained and lack adaptability when dealing with different tasks. In our work, we seek another way to use the semantic information, that is semantic-aware feature representation learning framework.Based on this, we propose SRMatcher, a new detector-free feature matching method, which encourages the network to learn integrated semantic feature representation.Specifically, to capture precise and rich semantics, we leverage the capabilities of recently popularized vision foundation models (VFMs) trained on extensive datasets. Then, a cross-images Semantic-aware Fusion Block (SFB) is proposed to integrate its fine-grained semantic features into the feature representation space. In this way, by reducing errors stemming from semantic inconsistencies in matching pairs, our proposed SRMatcher is able to deliver more accurate and realistic outcomes. Extensive experiments show that SRMatcher surpasses solid baselines and attains SOTA results on multiple real-world datasets. Compared to the previous SOTA approach GeoFormer, SRMatcher increases the area under the cumulative curve (AUC) by about 11% on HPatches. Additionally, the SRMatcher could serve as a plug-and-play framework for other matching methods like LoFTR, yielding substantial precision improvement., Comment: Accepted by ACM Multimedia 2024

Scanning electron microscopy (SEM) has been widely utilized in the field of materials science due to its significant advantages, such as large depth of field, wide field of view, and excellent stereoscopic imaging. However, at high magnification, the limited imaging range in SEM cannot cover all the possible inhomogeneous microstructures. In this research, we propose a novel approach for generating high-resolution SEM images across multiple scales, enabling a single image to capture physical dimensions at the centimeter level while preserving submicron-level details. We adopted the SEM imaging on the AlCoCrFeNi2.1 eutectic high entropy alloy (EHEA) as an example. SEM videos and image stitching are combined to fulfill this goal, and the video-extracted low-definition (LD) images are clarified by a well-trained denoising model. Furthermore, we segment the macroscopic image of the EHEA, and area of various microstructures are distinguished. Combining the segmentation results and hardness experiments, we found that the hardness is positively correlated with the content of body-centered cubic (BCC) phase, negatively correlated with the lamella width, and the relationship with the proportion of lamellar structures was not significant. Our work provides a feasible solution to generate macroscopic images based on SEMs for further analysis of the correlations between the microstructures and spatial distribution, and can be widely applied to other types of microscope., Comment: 22 pages,12 figures

While existing alignment paradigms have been integral in developing large language models (LLMs), LLMs often learn an averaged human preference and struggle to model diverse preferences across cultures, demographics, and communities. We propose Modular Pluralism, a modular framework based on multi-LLM collaboration for pluralistic alignment: it "plugs into" a base LLM a pool of smaller but specialized community LMs, where models collaborate in distinct modes to flexibility support three modes of pluralism: Overton, steerable, and distributional. Modular Pluralism is uniquely compatible with black-box LLMs and offers the modular control of adding new community LMs for previously underrepresented communities. We evaluate Modular Pluralism with six tasks and four datasets featuring questions/instructions with value-laden and perspective-informed responses. Extensive experiments demonstrate that Modular Pluralism advances the three pluralism objectives across six black-box and open-source LLMs. Further analysis reveals that LLMs are generally faithful to the inputs from smaller community LLMs, allowing seamless patching by adding a new community LM to better cover previously underrepresented communities., Comment: EMNLP 2024

Vision Language Models (VLMs) have received significant attention in recent years in the robotics community. VLMs are shown to be able to perform complex visual reasoning and scene understanding tasks, which makes them regarded as a potential universal solution for general robotics problems such as manipulation and navigation. However, previous VLMs for robotics such as RT-1, RT-2, and ManipLLM have focused on directly learning robot-centric actions. Such approaches require collecting a significant amount of robot interaction data, which is extremely costly in the real world. Thus, we propose A3VLM, an object-centric, actionable, articulation-aware vision language model. A3VLM focuses on the articulation structure and action affordances of objects. Its representation is robot-agnostic and can be translated into robot actions using simple action primitives. Extensive experiments in both simulation benchmarks and real-world settings demonstrate the effectiveness and stability of A3VLM. We release our code and other materials at https://github.com/changhaonan/A3VLM.

Learning general-purpose models from diverse datasets has achieved great success in machine learning. In robotics, however, existing methods in multi-task learning are typically constrained to a single robot and workspace, while recent work such as RT-X requires a non-trivial action normalization procedure to manually bridge the gap between different action spaces in diverse environments. In this paper, we propose the visual kinematics chain as a precise and universal representation of quasi-static actions for robot learning over diverse environments, which requires no manual adjustment since the visual kinematic chains can be automatically obtained from the robot's model and camera parameters. We propose the Visual Kinematics Transformer (VKT), a convolution-free architecture that supports an arbitrary number of camera viewpoints, and that is trained with a single objective of forecasting kinematic structures through optimal point-set matching. We demonstrate the superior performance of VKT over BC transformers as a general agent on Calvin, RLBench, Open-X, and real robot manipulation tasks. Video demonstrations can be found at https://mlzxy.github.io/visual-kinetic-chain., Comment: CoRL 2024

Differentially Private Stochastic Gradients Descent (DP-SGD) is a prominent paradigm for preserving privacy in deep learning. It ensures privacy by perturbing gradients with random noise calibrated to their entire norm at each training step. However, this perturbation suffers from a sub-optimal performance: it repeatedly wastes privacy budget on the general converging direction shared among gradients from different batches, which we refer as common knowledge, yet yields little information gain. Motivated by this, we propose a differentially private training framework with early gradient decomposition and reconstruction (DPDR), which enables more efficient use of the privacy budget. In essence, it boosts model utility by focusing on incremental information protection and recycling the privatized common knowledge learned from previous gradients at early training steps. Concretely, DPDR incorporates three steps. First, it disentangles common knowledge and incremental information in current gradients by decomposing them based on previous noisy gradients. Second, most privacy budget is spent on protecting incremental information for higher information gain. Third, the model is updated with the gradient reconstructed from recycled common knowledge and noisy incremental information. Theoretical analysis and extensive experiments show that DPDR outperforms state-of-the-art baselines on both convergence rate and accuracy., Comment: 14 pages

Large language models (LLMs) often incorporate multiple text chunks in their inputs to provide the necessary contexts. To speed up the prefill of the long LLM inputs, one can pre-compute the KV cache of a text and re-use the KV cache when the context is reused as the prefix of another LLM input. However, the reused text chunks are not always the input prefix, and when they are not, their precomputed KV caches cannot be directly used since they ignore the text's cross-attention with the preceding text in the LLM input. Thus, the benefits of reusing KV caches remain largely unrealized. This paper tackles just one question: when an LLM input contains multiple text chunks, how to quickly combine their precomputed KV caches in order to achieve the same generation quality as the expensive full prefill (i.e., without reusing KV cache)? We present CacheBlend, a scheme that reuses the pre-computed KV caches, regardless prefix or not, and selectively recomputes the KV values of a small subset of tokens to partially update each reused KV cache. In the meantime,the small extra delay for recomputing some tokens can be pipelined with the retrieval of KV caches within the same job,allowing CacheBlend to store KV caches in slower devices with more storage capacity while retrieving them without increasing the inference delay. By comparing CacheBlend with the state-of-the-art KV cache reusing schemes on three open-source LLMs of various sizes and four popular benchmark datasets of different tasks, we show that CacheBlend reduces time-to-first-token (TTFT) by 2.2-3.3X and increases the inference throughput by 2.8-5X, compared with full KV recompute, without compromising generation quality or incurring more storage cost.

Physics-guided neural networks (PGNN) is an effective tool that combines the benefits of data-driven modeling with the interpretability and generalization of underlying physical information. However, for a classical PGNN, the penalization of the physics-guided part is at the output level, which leads to a conservative result as systems with highly similar state-transition functions, i.e. only slight differences in parameters, can have significantly different time-series outputs. Furthermore, the classical PGNN cost function regularizes the model estimate over the entire state space with a constant trade-off hyperparameter. In this paper, we introduce a novel model augmentation strategy for nonlinear state-space model identification based on PGNN, using a weighted function regularization (W-PGNN). The proposed approach can efficiently augment the prior physics-based state-space models based on measurement data. A new weighted regularization term is added to the cost function to penalize the difference between the state and output function of the baseline physics-based and final identified model. This ensures the estimated model follows the baseline physics model functions in regions where the data has low information content, while placing greater trust in the data when a high informativity is present. The effectiveness of the proposed strategy over the current PGNN method is demonstrated on a benchmark example.

Children develop narrative skills by understanding and actively building connections between elements, image-text matching and consequences. However, it is challenging for children to clearly grasp these multi-level links only through explanations of text or facilitator's speech. To address this, we developed Colin, an interactive storytelling tool that supports children's multi-level narrative skills through both voice and visual modalities. In the generation stage, Colin supports facilitator to define and review generated text and image content freely. In the understanding stage, a question-feedback model helps children understand multi-level connections while co-creating stories with Colin. In the building phase, Colin actively encourages children to create connections between elements through drawing and speaking. A user study with 20 participants evaluated Colin by measuring children's engagement, understanding of cause-and-effect relationships, and the quality of their new story creations. Our results demonstrated that Colin significantly enhances the development of children's narrative skills across multiple levels.

In the realm of time series analysis, accurately measuring similarity is crucial for applications such as forecasting, anomaly detection, and clustering. However, existing metrics often fail to capture the complex, multidimensional nature of time series data, limiting their effectiveness and application. This paper introduces the Structured Similarity Index Measure for Time Series (TS3IM), a novel approach inspired by the success of the Structural Similarity Index Measure (SSIM) in image analysis, tailored to address these limitations by assessing structural similarity in time series. TS3IM evaluates multiple dimensions of similarity-trend, variability, and structural integrity-offering a more nuanced and comprehensive measure. This metric represents a significant leap forward, providing a robust tool for analyzing temporal data and offering more accurate and comprehensive sequence analysis and decision support in fields such as monitoring power consumption, analyzing traffic flow, and adversarial recognition. Our extensive experimental results also show that compared with traditional methods that rely heavily on computational correlation, TS3IM is 1.87 times more similar to Dynamic Time Warping (DTW) in evaluation results and improves by more than 50% in adversarial recognition., Comment: 6 pages, 6 figures

Video frame interpolation, the process of synthesizing intermediate frames between sequential video frames, has made remarkable progress with the use of event cameras. These sensors, with microsecond-level temporal resolution, fill information gaps between frames by providing precise motion cues. However, contemporary Event-Based Video Frame Interpolation (E-VFI) techniques often neglect the fact that event data primarily supply high-confidence features at scene edges during multi-modal feature fusion, thereby diminishing the role of event signals in optical flow (OF) estimation and warping refinement. To address this overlooked aspect, we introduce an end-to-end E-VFI learning method (referred to as EGMR) to efficiently utilize edge features from event signals for motion flow and warping enhancement. Our method incorporates an Edge Guided Attentive (EGA) module, which rectifies estimated video motion through attentive aggregation based on the local correlation of multi-modal features in a coarse-to-fine strategy. Moreover, given that event data can provide accurate visual references at scene edges between consecutive frames, we introduce a learned visibility map derived from event data to adaptively mitigate the occlusion problem in the warping refinement process. Extensive experiments on both synthetic and real datasets show the effectiveness of the proposed approach, demonstrating its potential for higher quality video frame interpolation.

Influencer selection in marketing involves choosing users with a strong online presence to promote products or services, leveraging their credibility and audience reach. This process is vital for its direct impact on brand visibility, consumer trust, and ultimately, sales conversion. Current research simplifies complex elements like user attitudes, thought processes, and advertising content into numerical values. This kind of approach risks missing the dynamic and contextual nuances crucial for developing effective influencer marketing strategies. To bridge this gap, we introduce a text-rich large Social Advertisement Graph (SAGraph) dataset collected from Weibo, a real-world influencer advertising platform. Our dataset centers around the advertising campaign for 6 products, consisting of 317,287 users, each with their profile information, and interaction data including 891,834 comments and 441,836 reposts. By leveraging this rich interaction and textual content, one can gain deeper insights into consumer behavior, refine influencer selection criteria, and develop more targeted and effective marketing strategies. We evaluated existing influencer selection baselines and the latest LLMs on this dataset, demonstrating the importance of textual content in advertising campaigns, as well as the availability and significant potential of LLMs for enhancing advertising strategies. We hope that this dataset will inspire further research: \url{https://github.com/xiaoqzhwhu/SAGraph/}.

In the digital era, the rapid propagation of fake news and rumors via social networks brings notable societal challenges and impacts public opinion regulation. Traditional fake news modeling typically forecasts the general popularity trends of different groups or numerically represents opinions shift. However, these methods often oversimplify real-world complexities and overlook the rich semantic information of news text. The advent of large language models (LLMs) provides the possibility of modeling subtle dynamics of opinion. Consequently, in this work, we introduce a Fake news Propagation Simulation framework (FPS) based on LLM, which studies the trends and control of fake news propagation in detail. Specifically, each agent in the simulation represents an individual with a distinct personality. They are equipped with both short-term and long-term memory, as well as a reflective mechanism to mimic human-like thinking. Every day, they engage in random opinion exchanges, reflect on their thinking, and update their opinions. Our simulation results uncover patterns in fake news propagation related to topic relevance, and individual traits, aligning with real-world observations. Additionally, we evaluate various intervention strategies and demonstrate that early and appropriately frequent interventions strike a balance between governance cost and effectiveness, offering valuable insights for practical applications. Our study underscores the significant utility and potential of LLMs in combating fake news.

Online movie review websites are valuable for information and discussion about movies. However, the massive spoiler reviews detract from the movie-watching experience, making spoiler detection an important task. Previous methods simply focus on reviews' text content, ignoring the heterogeneity of information in the platform. For instance, the metadata and the corresponding user's information of a review could be helpful. Besides, the spoiler language of movie reviews tends to be genre-specific, thus posing a domain generalization challenge for existing methods. To this end, we propose MMoE, a multi-modal network that utilizes information from multiple modalities to facilitate robust spoiler detection and adopts Mixture-of-Experts to enhance domain generalization. MMoE first extracts graph, text, and meta feature from the user-movie network, the review's textual content, and the review's metadata respectively. To handle genre-specific spoilers, we then adopt Mixture-of-Experts architecture to process information in three modalities to promote robustness. Finally, we use an expert fusion layer to integrate the features from different perspectives and make predictions based on the fused embedding. Experiments demonstrate that MMoE achieves state-of-the-art performance on two widely-used spoiler detection datasets, surpassing previous SOTA methods by 2.56% and 8.41% in terms of accuracy and F1-score. Further experiments also demonstrate MMoE's superiority in robustness and generalization.

Beyond the well-known giants like Uber Eats and DoorDash, there are hundreds of independent food delivery platforms in the United States. However, little is known about the sociotechnical landscape of these ``indie'' platforms. In this paper, we analyzed these platforms to understand why they were created, how they operate, and what technologies they use. We collected data on 495 indie platforms and detailed survey responses from 29 platforms. We found that personalized, timely service is a central value of indie platforms, as is a sense of responsibility to the local community they serve. Indie platforms are motivated to provide fair rates for restaurants and couriers. These alternative business practices differentiate them from mainstream platforms. Though indie platforms have plans to expand, a lack of customizability in off-the-shelf software prevents independent platforms from personalizing services for their local communities. We show that these platforms are a widespread and longstanding fixture of the food delivery market. We illustrate the diversity of motivations and values to explain why a one-size-fits-all support is insufficient, and we discuss the siloing of technology that inhibits platforms' growth. Through these insights, we aim to promote future HCI research into the potential development of public-interest technologies for local food delivery., Comment: To appear in CSCW 2024

To render each generated token in real-time for users, the Large Language Model (LLM) server generates tokens one by one and streams each token (or group of a few tokens) through the network to the user right after generation, which we refer to as LLM token streaming. However, under unstable network conditions, the LLM token streaming experience could suffer greatly from stalls since one packet loss could block the rendering of later tokens even if the packets containing them arrive on time. With a measurement study, we show that current applications suffer from increased stalls under unstable networks. For this emerging token streaming problem in LLM Chatbots that differs from previous multimedia and text applications, we propose a novel transmission scheme, called Eloquent, which puts newly generated tokens as well as currently unacknowledged tokens in the next outgoing packet. This ensures that each packet contains some new tokens and, in the meantime, is independently rendered when received, avoiding the aforementioned stalls caused by missing packets. Through simulation under various networks, we show Eloquent reduces stall ratio (proportion of token rendering wait time) by 71.0% compared to the retransmission method commonly used by real chatbot applications and by 31.6% compared to the baseline packet duplication scheme. By tailoring Eloquent to fit the token-by-token generation of LLM, we enable the Chatbots to respond like an eloquent speaker for users to better enjoy pervasive AI., Comment: In SIGCOMM Workshop on Networks for AI Computing (NAIC '24)

Training networks consisting of biophysically accurate neuron models could allow for new insights into how brain circuits can organize and solve tasks. We begin by analyzing the extent to which the central algorithm for neural network learning -- stochastic gradient descent through backpropagation (BP) -- can be used to train such networks. We find that properties of biophysically based neural network models needed for accurate modelling such as stiffness, high nonlinearity and long evaluation timeframes relative to spike times makes BP unstable and divergent in a variety of cases. To address these instabilities and inspired by recent work, we investigate the use of "gradient-estimating" evolutionary algorithms (EAs) for training biophysically based neural networks. We find that EAs have several advantages making them desirable over direct BP, including being forward-pass only, robust to noisy and rigid losses, allowing for discrete loss formulations, and potentially facilitating a more global exploration of parameters. We apply our method to train a recurrent network of Morris-Lecar neuron models on a stimulus integration and working memory task, and show how it can succeed in cases where direct BP is inapplicable. To expand on the viability of EAs in general, we apply them to a general neural ODE problem and a stiff neural ODE benchmark and find again that EAs can out-perform direct BP here, especially for the over-parameterized regime. Our findings suggest that biophysical neurons could provide useful benchmarks for testing the limits of BP-adjacent methods, and demonstrate the viability of EAs for training networks with complex components.

In this work, we take a first step towards designing summarization systems that are faithful to the author's intent, not only the semantic content of the article. Focusing on a case study of preserving political perspectives in news summarization, we find that existing approaches alter the political opinions and stances of news articles in more than 50% of summaries, misrepresenting the intent and perspectives of the news authors. We thus propose P^3SUM, a diffusion model-based summarization approach controlled by political perspective classifiers. In P^3SUM, the political leaning of a generated summary is iteratively evaluated at each decoding step, and any drift from the article's original stance incurs a loss back-propagated to the embedding layers, steering the political stance of the summary at inference time. Extensive experiments on three news summarization datasets demonstrate that P^3SUM outperforms state-of-the-art summarization systems and large language models by up to 13.7% in terms of the success rate of stance preservation, with competitive performance on standard metrics of summarization quality. Our findings present a first analysis of preservation of pragmatic features in summarization, highlight the lacunae in existing summarization models -- that even state-of-the-art models often struggle to preserve author's intents -- and develop new summarization systems that are more faithful to author's perspectives.

One major challenge for autonomous attitude takeover control for on-orbit servicing of spacecraft is that an accurate dynamic motion model of the combined vehicles is highly nonlinear, complex and often costly to identify online, which makes traditional model-based control impractical for this task. To address this issue, a recursive online sparse Gaussian Process (GP)-based learning strategy for attitude takeover control of noncooperative targets with maneuverability is proposed, where the unknown dynamics are online compensated based on the learnt GP model in a semi-feedforward manner. The method enables the continuous use of on-orbit data to successively improve the learnt model during online operation and has reduced computational load compared to standard GP regression. Next to the GP-based feedforward, a feedback controller is proposed that varies its gains based on the predicted model confidence, ensuring robustness of the overall scheme. Moreover, rigorous theoretical proofs of Lyapunov stability and boundedness guarantees of the proposed method-driven closed-loop system are provided in the probabilistic sense. A simulation study based on a high-fidelity simulator is used to show the effectiveness of the proposed strategy and demonstrate its high performance., Comment: 17 pages, 14 figures. Submitted to in IEEE Transactions on Aerospace and Electronic Systems

In theoretical neuroscience, recent work leverages deep learning tools to explore how some network attributes critically influence its learning dynamics. Notably, initial weight distributions with small (resp. large) variance may yield a rich (resp. lazy) regime, where significant (resp. minor) changes to network states and representation are observed over the course of learning. However, in biology, neural circuit connectivity could exhibit a low-rank structure and therefore differs markedly from the random initializations generally used for these studies. As such, here we investigate how the structure of the initial weights -- in particular their effective rank -- influences the network learning regime. Through both empirical and theoretical analyses, we discover that high-rank initializations typically yield smaller network changes indicative of lazier learning, a finding we also confirm with experimentally-driven initial connectivity in recurrent neural networks. Conversely, low-rank initialization biases learning towards richer learning. Importantly, however, as an exception to this rule, we find lazier learning can still occur with a low-rank initialization that aligns with task and data statistics. Our research highlights the pivotal role of initial weight structures in shaping learning regimes, with implications for metabolic costs of plasticity and risks of catastrophic forgetting., Comment: Published at ICLR 2024

As large language models (LLMs) take on complex tasks, their inputs are supplemented with longer contexts that incorporate domain knowledge. Yet using long contexts is challenging, as nothing can be generated until the whole context is processed by the LLM. While the context-processing delay can be reduced by reusing the KV cache of a context across different inputs, fetching the KV cache, which contains large tensors, over the network can cause high extra network delays. CacheGen is a fast context-loading module for LLM systems. First, CacheGen uses a custom tensor encoder, leveraging KV cache's distributional properties to encode a KV cache into more compact bitstream representations with negligible decoding overhead, to save bandwidth usage. Second, CacheGen adapts the compression level of different parts of a KV cache to cope with changes in available bandwidth, in order to maintain low context-loading delay and high generation quality. % When available bandwidth drops, CacheGen may raise the compression level for a part of the context or recompute its KV cache on the fly. We test CacheGen on popular LLMs and datasets. Compared to the recent systems that reuse the KV cache, CacheGen reduces the KV cache size by 3.5-4.3x and the total delay in fetching and processing contexts by 3.2-3.7x with negligible impact on the LLM response quality. Our code is at: https://github.com/UChi-JCL/CacheGen., Comment: SIGCOMM'24

ML APIs have greatly relieved application developers of the burden to design and train their own neural network models -- classifying objects in an image can now be as simple as one line of Python code to call an API. However, these APIs offer the same pre-trained models regardless of how their output is used by different applications. This can be suboptimal as not all ML inference errors can cause application failures, and the distinction between inference errors that can or cannot cause failures varies greatly across applications. To tackle this problem, we first study 77 real-world applications, which collectively use six ML APIs from two providers, to reveal common patterns of how ML API output affects applications' decision processes. Inspired by the findings, we propose ChameleonAPI, an optimization framework for ML APIs, which takes effect without changing the application source code. ChameleonAPI provides application developers with a parser that automatically analyzes the application to produce an abstract of its decision process, which is then used to devise an application-specific loss function that only penalizes API output errors critical to the application. ChameleonAPI uses the loss function to efficiently train a neural network model customized for each application and deploys it to serve API invocations from the respective application via existing interface. Compared to a baseline that selects the best-of-all commercial ML API, we show that ChameleonAPI reduces incorrect application decisions by 43%.

Deep learning inference on streaming media data, such as object detection in video or LiDAR feeds and text extraction from audio waves, is now ubiquitous. To achieve high inference accuracy, these applications typically require significant network bandwidth to gather high-fidelity data and extensive GPU resources to run deep neural networks (DNNs). While the high demand for network bandwidth and GPU resources could be substantially reduced by optimally adapting the configuration knobs, such as video resolution and frame rate, current adaptation techniques fail to meet three requirements simultaneously: adapt configurations (i) with minimum extra GPU or bandwidth overhead; (ii) to reach near-optimal decisions based on how the data affects the final DNN's accuracy, and (iii) do so for a range of configuration knobs. This paper presents OneAdapt, which meets these requirements by leveraging a gradient-ascent strategy to adapt configuration knobs. The key idea is to embrace DNNs' differentiability to quickly estimate the accuracy's gradient to each configuration knob, called AccGrad. Specifically, OneAdapt estimates AccGrad by multiplying two gradients: InputGrad (i.e. how each configuration knob affects the input to the DNN) and DNNGrad (i.e. how the DNN input affects the DNN inference output). We evaluate OneAdapt across five types of configurations, four analytic tasks, and five types of input data. Compared to state-of-the-art adaptation schemes, OneAdapt cuts bandwidth usage and GPU usage by 15-59% while maintaining comparable accuracy or improves accuracy by 1-5% while using equal or fewer resources., Comment: SoCC' 23

We propose Knowledge Crosswords, a geometric knowledge reasoning benchmark consisting of incomplete knowledge networks bounded by structured factual constraints, where LLMs are tasked with inferring the missing facts to meet all constraints. The novel setting of geometric knowledge reasoning necessitates new LM abilities beyond existing atomic/linear multi-hop QA, such as backtracking, verifying facts and constraints, reasoning with uncertainty, and more. Knowledge Crosswords contains 2,101 individual problems, covering diverse knowledge domains, and is further divided into three difficulty levels. We conduct extensive experiments to evaluate existing LLMs and approaches on Knowledge Crosswords. Results demonstrate that baseline approaches struggle with larger knowledge networks and semantically-equivalent entity distractors. In light of their limitations, we propose two new approaches, Staged Prompting and Verify-All, to augment LLMs' abilities for error-aware backtracking and constraint verification. Our Verify-All significantly outperforms prior methods and is more robust towards problems in the hard subset. Further analysis shows that geometric knowledge reasoning poses new challenges to LLMs' knowledge abilities, particularly in robustness towards varying option orders, complex structural constraints in knowledge networks, "none of the above" scenarios, and more.