arXiv subjects

Feng Chen

Publications and source records attributed to Feng Chen.

At least 181 records · Page 10Linked to original sources

arXiv2024

ZipCache: A DRAM/SSD Cache with Built-in Transparent Compression

As a core component in modern data centers, key-value cache provides high-throughput and low-latency services for high-speed data processing. The effectiveness of a key-value cache relies on its ability of accommodating the needed data. However, expanding the cache capacity is often more difficult than commonly expected because of many practical constraints, such as server costs, cooling issues, rack space, and even human resource expenses. A potential solution is compression, which virtually extends the cache capacity by condensing data in cache. In practice, this seemingly simple idea has not gained much traction in key-value cache system design, due to several critical issues: the compression-unfriendly index structure, severe read/write amplification, wasteful decompression operations, and heavy computing cost. This paper presents a hybrid DRAM-SSD cache design to realize a systematic integration of data compression in key-value cache. By treating compression as an essential component, we have redesigned the indexing structure, data management, and leveraged the emerging computational SSD hardware for collaborative optimizations. We have developed a prototype, called ZipCache. Our experimental results show that ZipCache can achieve up to 72.4% higher throughput and 42.4% lower latency, while reducing the write amplification by up to 26.2 times.

cs.DB↗

arXiv2024

Stable Continual Reinforcement Learning via Diffusion-based Trajectory Replay

Given the inherent non-stationarity prevalent in real-world applications, continual Reinforcement Learning (RL) aims to equip the agent with the capability to address a series of sequentially presented decision-making tasks. Within this problem setting, a pivotal challenge revolves around \textit{catastrophic forgetting} issue, wherein the agent is prone to effortlessly erode the decisional knowledge associated with past encountered tasks when learning the new one. In recent progresses, the \textit{generative replay} methods have showcased substantial potential by employing generative models to replay data distribution of past tasks. Compared to storing the data from past tasks directly, this category of methods circumvents the growing storage overhead and possible data privacy concerns. However, constrained by the expressive capacity of generative models, existing \textit{generative replay} methods face challenges in faithfully reconstructing the data distribution of past tasks, particularly in scenarios with a myriad of tasks or high-dimensional data. Inspired by the success of diffusion models in various generative tasks, this paper introduces a novel continual RL algorithm DISTR (Diffusion-based Trajectory Replay) that employs a diffusion model to memorize the high-return trajectory distribution of each encountered task and wakeups these distributions during the policy learning on new tasks. Besides, considering the impracticality of replaying all past data each time, a prioritization mechanism is proposed to prioritize the trajectory replay of pivotal tasks in our method. Empirical experiments on the popular continual RL benchmark \texttt{Continual World} demonstrate that our proposed method obtains a favorable balance between \textit{stability} and \textit{plasticity}, surpassing various existing continual RL baselines in average success rate.

cs.LG↗

arXiv2024

Feature-Space Semantic Invariance: Enhanced OOD Detection for Open-Set Domain Generalization

Open-set domain generalization addresses a real-world challenge: training a model to generalize across unseen domains (domain generalization) while also detecting samples from unknown classes not encountered during training (open-set recognition). However, most existing approaches tackle these issues separately, limiting their practical applicability. To overcome this limitation, we propose a unified framework for open-set domain generalization by introducing Feature-space Semantic Invariance (FSI). FSI maintains semantic consistency across different domains within the feature space, enabling more accurate detection of OOD instances in unseen domains. Additionally, we adopt a generative model to produce synthetic data with novel domain styles or class labels, enhancing model robustness. Initial experiments show that our method improves AUROC by 9.1% to 18.9% on ColoredMNIST, while also significantly increasing in-distribution classification accuracy.

cs.CV↗

arXiv2024

FEED: Fairness-Enhanced Meta-Learning for Domain Generalization

Generalizing to out-of-distribution data while being aware of model fairness is a significant and challenging problem in meta-learning. The goal of this problem is to find a set of fairness-aware invariant parameters of classifier that is trained using data drawn from a family of related training domains with distribution shift on non-sensitive features as well as different levels of dependence between model predictions and sensitive features so that the classifier can achieve good generalization performance on unknown but distinct test domains. To tackle this challenge, existing state-of-the-art methods either address the domain generalization problem but completely ignore learning with fairness or solely specify shifted domains with various fairness levels. This paper introduces an approach to fairness-aware meta-learning that significantly enhances domain generalization capabilities. Our framework, Fairness-Enhanced Meta-Learning for Domain Generalization (FEED), disentangles latent data representations into content, style, and sensitive vectors. This disentanglement facilitates the robust generalization of machine learning models across diverse domains while adhering to fairness constraints. Unlike traditional methods that focus primarily on domain invariance or sensitivity to shifts, our model integrates a fairness-aware invariance criterion directly into the meta-learning process. This integration ensures that the learned parameters uphold fairness consistently, even when domain characteristics vary widely. We validate our approach through extensive experiments across multiple benchmarks, demonstrating not only superior performance in maintaining high accuracy and fairness but also significant improvements over existing state-of-the-art methods in domain generalization tasks.

cs.LG↗

arXiv2024

MADOD: Generalizing OOD Detection to Unseen Domains via G-Invariance Meta-Learning

Real-world machine learning applications often face simultaneous covariate and semantic shifts, challenging traditional domain generalization and out-of-distribution (OOD) detection methods. We introduce Meta-learned Across Domain Out-of-distribution Detection (MADOD), a novel framework designed to address both shifts concurrently. MADOD leverages meta-learning and G-invariance to enhance model generalizability and OOD detection in unseen domains. Our key innovation lies in task construction: we randomly designate in-distribution classes as pseudo-OODs within each meta-learning task, simulating OOD scenarios using existing data. This approach, combined with energy-based regularization, enables the learning of robust, domain-invariant features while calibrating decision boundaries for effective OOD detection. Operating in a test domain-agnostic setting, MADOD eliminates the need for adaptation during inference, making it suitable for scenarios where test data is unavailable. Extensive experiments on real-world and synthetic datasets demonstrate MADOD's superior performance in semantic OOD detection across unseen domains, achieving an AUPR improvement of 8.48% to 20.81%, while maintaining competitive in-distribution classification accuracy, representing a significant advancement in handling both covariate and semantic shifts.

cs.LG↗

arXiv2024

Scenario Predict-then-Optimize for Data-Driven Online Inventory Routing

The real-time joint optimization of inventory replenishment and vehicle routing is essential for cost-efficiently operating one-warehouse, multiple-retailer systems. This is complex, as future demand predictions should capture correlation retailer demand, and based upon such predictions, replenishment and routing decisions must be taken. Traditionally, such decisions are made by either making distributional assumptions or using machine-learning-based point forecasts. The former approach ignores nonstationary demand patterns, while the latter approach only provides a point forecast ignoring the inherent forecast error. Consequently, in practice, service levels often do not meet their targets, and truck fill rates fall short, harming the efficiency and sustainability of daily operations. We propose Scenario Predict-then-Optimize. This fully data-driven approach for online inventory routing consists of two subsequent steps at each real-time decision epoch. The scenario-predict step exploits neural networks, specifically multi-horizon quantile recurrent neural networks, to predict future demand quantiles, upon which we design a scenario sampling approach. The subsequent scenario-optimize step then solves a scenario-based stochastic programming approximation. Results show that our approach outperforms the classic Predict-then-Optimize paradigm and empirical sampling methods. We show this both on synthetic data and real-life data. Our approach is appealing to practitioners. It is fast, does not rely on distributional assumptions, and does not face the burden of single-scenario forecasts. We show it is robust for various demand and cost parameters, enhancing the efficiency and sustainability of daily replenishment and routing decisions. Finally, Scenario Predict-then-Optimize is general and can be easily extended to account for other operational constraints, making it a useful tool in practice.

math.OC↗

arXiv2024

Rethinking the Uncertainty: A Critical Review and Analysis in the Era of Large Language Models

In recent years, Large Language Models (LLMs) have become fundamental to a broad spectrum of artificial intelligence applications. As the use of LLMs expands, precisely estimating the uncertainty in their predictions has become crucial. Current methods often struggle to accurately identify, measure, and address the true uncertainty, with many focusing primarily on estimating model confidence. This discrepancy is largely due to an incomplete understanding of where, when, and how uncertainties are injected into models. This paper introduces a comprehensive framework specifically designed to identify and understand the types and sources of uncertainty, aligned with the unique characteristics of LLMs. Our framework enhances the understanding of the diverse landscape of uncertainties by systematically categorizing and defining each type, establishing a solid foundation for developing targeted methods that can precisely quantify these uncertainties. We also provide a detailed introduction to key related concepts and examine the limitations of current methods in mission-critical and safety-sensitive applications. The paper concludes with a perspective on future directions aimed at enhancing the reliability and practical adoption of these methods in real-world scenarios.

cs.AI↗

arXiv2024

Integrated spectrally multiplexed light-matter interface at telecom band

Light-matter interface is an important building block for long-distance quantum networks. Towards a scalable quantum network with high-rate quantum information processing, it requires to develop integrated light-matter interfaces with broadband and multiplexing capacities. Here we demonstrate a light-matter interface at telecom band in an integrated system. A five-spectral-channel atomic-frequency-comb photonic memory is prepared on a laser-written Er3+:LiNbO3 chip. The bandwidth of each channel is 4 GHz with a channel spacing of 15 GHz. The signal photons from time-bin entangled photon pairs at telecom band are sent into the on-chip memory and recalled after a storage time of 152 ns. The entanglement-preserving nature of our integrated quantum interface is assessed by an input/output fidelity of >92% for all the five spectral channels. Our light-matter interfaces constitute a notable step forward toward a high-rate quantum network involving integrated device.

quant-ph↗

arXiv2024

Get rich quick: exact solutions reveal how unbalanced initializations promote rapid feature learning

While the impressive performance of modern neural networks is often attributed to their capacity to efficiently extract task-relevant features from data, the mechanisms underlying this rich feature learning regime remain elusive, with much of our theoretical understanding stemming from the opposing lazy regime. In this work, we derive exact solutions to a minimal model that transitions between lazy and rich learning, precisely elucidating how unbalanced layer-specific initialization variances and learning rates determine the degree of feature learning. Our analysis reveals that they conspire to influence the learning regime through a set of conserved quantities that constrain and modify the geometry of learning trajectories in parameter and function space. We extend our analysis to more complex linear models with multiple neurons, outputs, and layers and to shallow nonlinear networks with piecewise linear activation functions. In linear networks, rapid feature learning only occurs from balanced initializations, where all layers learn at similar speeds. While in nonlinear networks, unbalanced initializations that promote faster learning in earlier layers can accelerate rich learning. Through a series of experiments, we provide evidence that this unbalanced rich regime drives feature learning in deep finite-width networks, promotes interpretability of early layers in CNNs, reduces the sample complexity of learning hierarchical data, and decreases the time to grokking in modular arithmetic. Our theory motivates further exploration of unbalanced initializations to enhance efficient feature learning.

cs.LG↗

arXiv2024

On the Evaluation of Generative Robotic Simulations

Due to the difficulty of acquiring extensive real-world data, robot simulation has become crucial for parallel training and sim-to-real transfer, highlighting the importance of scalable simulated robotic tasks. Foundation models have demonstrated impressive capacities in autonomously generating feasible robotic tasks. However, this new paradigm underscores the challenge of adequately evaluating these autonomously generated tasks. To address this, we propose a comprehensive evaluation framework tailored to generative simulations. Our framework segments evaluation into three core aspects: quality, diversity, and generalization. For single-task quality, we evaluate the realism of the generated task and the completeness of the generated trajectories using large language models and vision-language models. In terms of diversity, we measure both task and data diversity through text similarity of task descriptions and world model loss trained on collected task trajectories. For task-level generalization, we assess the zero-shot generalization ability on unseen tasks of a policy trained with multiple generated tasks. Experiments conducted on three representative task generation pipelines demonstrate that the results from our framework are highly consistent with human evaluations, confirming the feasibility and validity of our approach. The findings reveal that while metrics of quality and diversity can be achieved through certain methods, no single approach excels across all metrics, suggesting a need for greater focus on balancing these different metrics. Additionally, our analysis further highlights the common challenge of low generalization capability faced by current works. Our anonymous website: https://sites.google.com/view/evaltasks.

cs.RO↗

arXiv2024

Can We Trust the Performance Evaluation of Uncertainty Estimation Methods in Text Summarization?

Text summarization, a key natural language generation (NLG) task, is vital in various domains. However, the high cost of inaccurate summaries in risk-critical applications, particularly those involving human-in-the-loop decision-making, raises concerns about the reliability of uncertainty estimation on text summarization (UE-TS) evaluation methods. This concern stems from the dependency of uncertainty model metrics on diverse and potentially conflicting NLG metrics. To address this issue, we introduce a comprehensive UE-TS benchmark incorporating 31 NLG metrics across four dimensions. The benchmark evaluates the uncertainty estimation capabilities of two large language models and one pre-trained language model on three datasets, with human-annotation analysis incorporated where applicable. We also assess the performance of 14 common uncertainty estimation methods within this benchmark. Our findings emphasize the importance of considering multiple uncorrelated NLG metrics and diverse uncertainty estimation methods to ensure reliable and efficient evaluation of UE-TS techniques. Our code and data are available https://github.com/he159ok/Benchmark-of-Uncertainty-Estimation-Methods-in-Text-Summarization.

cs.CL↗

arXiv2024

Hopping Transfer Optimizes Avalanche Multiplication in Molybdenum Disulfide

Recently, avalanche multiplication has been observed in TMDC-based FETs, enhancing sensor performance with high sensitivity. However, the high voltage required for operation can damage the FETs, making it crucial to reduce the breakdown voltage for effective sensing applications. Here, we demonstrate that the utilization of hopping transfer induced by high-density defects can effectively reduce the breakdown voltage in TMDCs FETs. By substituting oxygen atoms for sulfur atoms in a monolayer of MoS2, we create MoS2-xOx, with x carefully adjusted within the range of 0 to 0.51. Oxygen doping reduces the bandgap of TMDCs and enhances ion collision rates. Moreover, higher levels of oxygen doping (x > 0.41) in MoS2-xOx exhibit nearest-neighbor hopping behavior, leading to a significant enhancement in electron mobility. These improvements result in a decrease in the breakdown voltage of avalanche multiplication from 26.2 V to 12.6 V. Additionally, we propose avalanche multiplication in MoS2-xOx as an efficient sensing mechanism to overcome the limitations of gas sensing. The MoS2-xOx sensors display an ultra-high response to NO2 gas in the air, with a response of 5.8x103 % to NO2 gas of 50 ppb at room temperature, which is nearly two orders of magnitude higher than resistance-type gas detectors based on TMDCs. This work demonstrates that hopping transfer induced by high-density oxygen defects can effectively decrease the breakdown voltage of MoS2-xOx FETs, enhancing avalanche multiplication and serving as a promising mechanism for ultrasensitive gas detection.

physics.app-ph↗

arXiv2024

IAFI-FCOS: Intra- and across-layer feature interaction FCOS model for lesion detection of CT images

Effective lesion detection in medical image is not only rely on the features of lesion region,but also deeply relative to the surrounding information.However,most current methods have not fully utilize it.What is more,multi-scale feature fusion mechanism of most traditional detectors are unable to transmit detail information without loss,which makes it hard to detect small and boundary ambiguous lesion in early stage disease.To address the above issues,we propose a novel intra- and across-layer feature interaction FCOS model (IAFI-FCOS) with a multi-scale feature fusion mechanism ICAF-FPN,which is a network structure with intra-layer context augmentation (ICA) block and across-layer feature weighting (AFW) block.Therefore,the traditional FCOS detector is optimized by enriching the feature representation from two perspectives.Specifically,the ICA block utilizes dilated attention to augment the context information in order to capture long-range dependencies between the lesion region and the surrounding.The AFW block utilizes dual-axis attention mechanism and weighting operation to obtain the efficient across-layer interaction features,enhancing the representation of detailed features.Our approach has been extensively experimented on both the private pancreatic lesion dataset and the public DeepLesion dataset,our model achieves SOTA results on the pancreatic lesion dataset.

cs.CV↗

arXiv2024

Periodic Coronal Rain Driven by Self-consistent Heating Process in a Radiative Magnetohydrodynamic Simulation

The periodic coronal rain and in-phase radiative intensity pulsations have been observed in multiple wavelengths in recent years. However, due to the lack of three-dimensional coronal magnetic fields and thermodynamic data in observations, it remains challenging to quantify the coronal heating rate that drives the mass cycles. In this work, based on the MURaM code, we conduct a three-dimensional radiative magnetohydrodynamic simulation spanning from the convective zone to the corona, where the solar atmosphere is heated self-consistently through dissipation resulting from magneto-convection. For the first time, we model the periodic coronal rain in an active region. With a high spatial resolution, the simulation well resembles the observational features across different extreme ultraviolet wavelengths. These include the realistic interweaving coronal loops, periodic coronal rain and periodic intensity pulsations, with two periods of 3.0~h and 3.7~h identified within one loop system. Moreover, the simulation allows for a detailed three-dimensional depiction of coronal rain on small scales, revealing adjacent shower-like rain clumps $\sim500$~km in width and showcasing their multi-thermal internal structures. We further reveal that these periodic variations essentially reflect the cyclic energy evolution of the coronal loop under thermal non-equilibrium state. Importantly, as the driver of the mass circulation, the self-consistent coronal heating rate is considerably complex in time and space, with hour-level variations in one order of magnitude, minute-level bursts, and varying asymmetry reaching ten times between footpoints. This provides an instructive template for the ad hoc heating function, and further enhances our understanding of the coronal heating process.

astro-ph.SR↗

arXiv2024

LSM-YOLO: A Compact and Effective ROI Detector for Medical Detection

In existing medical Region of Interest (ROI) detection, there lacks an algorithm that can simultaneously satisfy both real-time performance and accuracy, not meeting the growing demand for automatic detection in medicine. Although the basic YOLO framework ensures real-time detection due to its fast speed, it still faces challenges in maintaining precision concurrently. To alleviate the above problems, we propose a novel model named Lightweight Shunt Matching-YOLO (LSM-YOLO), with Lightweight Adaptive Extraction (LAE) and Multipath Shunt Feature Matching (MSFM). Firstly, by using LAE to refine feature extraction, the model can obtain more contextual information and high-resolution details from multiscale feature maps, thereby extracting detailed features of ROI in medical images while reducing the influence of noise. Secondly, MSFM is utilized to further refine the fusion of high-level semantic features and low-level visual features, enabling better fusion between ROI features and neighboring features, thereby improving the detection rate for better diagnostic assistance. Experimental results demonstrate that LSM-YOLO achieves 48.6% AP on a private dataset of pancreatic tumors, 65.1% AP on the BCCD blood cell detection public dataset, and 73.0% AP on the Br35h brain tumor detection public dataset. Our model achieves state-of-the-art performance with minimal parameter cost on the above three datasets. The source codes are at: https://github.com/VincentYuuuuuu/LSM-YOLO.

cs.CV↗

arXiv2024

Learning Causally Disentangled Representations via the Principle of Independent Causal Mechanisms

Learning disentangled causal representations is a challenging problem that has gained significant attention recently due to its implications for extracting meaningful information for downstream tasks. In this work, we define a new notion of causal disentanglement from the perspective of independent causal mechanisms. We propose ICM-VAE, a framework for learning causally disentangled representations supervised by causally related observed labels. We model causal mechanisms using nonlinear learnable flow-based diffeomorphic functions to map noise variables to latent causal variables. Further, to promote the disentanglement of causal factors, we propose a causal disentanglement prior learned from auxiliary labels and the latent causal structure. We theoretically show the identifiability of causal factors and mechanisms up to permutation and elementwise reparameterization. We empirically demonstrate that our framework induces highly disentangled causal factors, improves interventional robustness, and is compatible with counterfactual generation.

cs.LG↗

arXiv2024

Causal Diffusion Autoencoders: Toward Counterfactual Generation via Diffusion Probabilistic Models

Diffusion probabilistic models (DPMs) have become the state-of-the-art in high-quality image generation. However, DPMs have an arbitrary noisy latent space with no interpretable or controllable semantics. Although there has been significant research effort to improve image sample quality, there is little work on representation-controlled generation using diffusion models. Specifically, causal modeling and controllable counterfactual generation using DPMs is an underexplored area. In this work, we propose CausalDiffAE, a diffusion-based causal representation learning framework to enable counterfactual generation according to a specified causal model. Our key idea is to use an encoder to extract high-level semantically meaningful causal variables from high-dimensional data and model stochastic variation using reverse diffusion. We propose a causal encoding mechanism that maps high-dimensional data to causally related latent factors and parameterize the causal mechanisms among latent factors using neural networks. To enforce the disentanglement of causal variables, we formulate a variational objective and leverage auxiliary label information in a prior to regularize the latent space. We propose a DDIM-based counterfactual generation procedure subject to do-interventions. Finally, to address the limited label supervision scenario, we also study the application of CausalDiffAE when a part of the training data is unlabeled, which also enables granular control over the strength of interventions in generating counterfactuals during inference. We empirically show that CausalDiffAE learns a disentangled latent space and is capable of generating high-quality counterfactual images.

cs.LG↗

arXiv2024

Likelihood inference of the non-stationary Hawkes process with non-exponential kernel

The Hawkes process is a popular point process model for event sequences that exhibit temporal clustering. The intensity process of a Hawkes process consists of two components, the baseline intensity and the accumulated excitation effect due to past events, with the latter specified via an excitation kernel. The classical Hawkes process assumes a constant baseline intensity and an exponential excitation kernel. This results in an intensity process that is Markovian, a fact that has been used extensively to establish the strong consistency and asymtpotic normality of maximum likelihood estimators or similar. However, these assumptions can be overly restrictive and unrealistic for modelling the many applications which require the baseline intensity to vary with time and the excitation kernel to have non-exponential decay. However, asymptotic properties of maximum likelihood inference for the parameters specifying the baseline intensity and the self-exciting decay under this setup are substantially more difficult since the resulting intensity process is non-Markovian. To overcome this challenge, we develop an approximation procedure to show the intensity process is asymptotically ergodic in a suitably defined sense. This allows for the identification of an ergodic limit to the likelihood function and its derivatives, as required for obtaining large sample inference under minimal regularity conditions.

math.ST↗