arXiv subjects

Rui Li

Publications and source records attributed to Rui Li.

At least 199 records · Page 11Linked to original sources

arXiv2025

Request-Only Optimization for Recommendation Systems

Deep Learning Recommendation Models (DLRMs) represent one of the largest machine learning applications on the planet. Industry-scale DLRMs are trained with petabytes of recommendation data to serve billions of users every day. To utilize the rich user signals in the long user history, DLRMs have been scaled up to unprecedented complexity, up to trillions of floating-point operations (TFLOPs) per example. This scale, coupled with the huge amount of training data, necessitates new storage and training algorithms to efficiently improve the quality of these complex recommendation systems. In this paper, we present a Request-Only Optimizations (ROO) training and modeling paradigm. ROO simultaneously improves the storage and training efficiency as well as the model quality of recommendation systems. We holistically approach this challenge through co-designing data (i.e., request-only data), infrastructure (i.e., request-only based data processing pipeline), and model architecture (i.e., request-only neural architectures). Our ROO training and modeling paradigm treats a user request as a unit of the training data. Compared with the established practice of treating a user impression as a unit, our new design achieves native feature deduplication in data logging, consequently saving data storage. Second, by de-duplicating computations and communications across multiple impressions in a request, this new paradigm enables highly scaled-up neural network architectures to better capture user interest signals, such as Generative Recommenders (GRs) and other request-only friendly architectures.

cs.IR↗

arXiv2025

CSST Strong Lensing Preparation: Fast Modeling of Galaxy-Galaxy Strong Lenses in the Big Data Era

Galaxy-galaxy strong lensing provides a powerful probe of galaxy formation, evolution, and the properties of dark matter and dark energy. However, conventional lens-modeling approaches are computationally expensive and require fine-tuning to avoid local optima, rendering them impractical for the hundreds of thousands of lenses expected from surveys such as Euclid, CSST, and Roman Space Telescopes. To overcome these challenges, we introduce TinyLensGPU, a GPU-accelerated lens-modeling tool that employs XLA-based acceleration with JAX and a neural-network-enhanced nested sampling algorithm, nautilus-sampler. Tests on 1,000 simulated galaxy-galaxy lenses demonstrate that on an RTX 4060 Ti GPU, TinyLensGPU achieves likelihood evaluations approximately 2,000 times faster than traditional methods. Moreover, the nautilus-sampler reduces the number of likelihood evaluations by a factor of 3, decreasing the overall modeling time per lens from several days to roughly 3 minutes. Application to 63 SLACS lenses observed by the Hubble Space Telescope recovers Einstein radii consistent with the literature values (within $\lesssim 5\%$ deviation), which is within known systematic uncertainties. Catastrophic failures, where the sampler becomes trapped in local optima, occur in approximately 5\% of the simulated cases and 10\% of the SLACS sample. We argue that such issues are inherent to automated lens modeling but can be mitigated by incorporating prior knowledge from machine learning techniques. This work thus marks a promising step toward the efficient analysis of strong lenses in the era of big data. The code and data are available online: https://github.com/caoxiaoyue/TinyLensGpu.

astro-ph.IM↗

arXiv2025

Fitness aligned structural modeling enables scalable virtual screening with AuroBind

Most human proteins remain undrugged, over 96% of human proteins remain unexploited by approved therapeutics. While structure-based virtual screening promises to expand the druggable proteome, existing methods lack atomic-level precision and fail to predict binding fitness, limiting translational impact. We present AuroBind, a scalable virtual screening framework that fine-tunes a custom atomic-level structural model on million-scale chemogenomic data. AuroBind integrates direct preference optimization, self-distillation from high-confidence complexes, and a teacher-student acceleration strategy to jointly predict ligand-bound structures and binding fitness. The proposed models outperform state-of-the-art models on structural and functional benchmarks while enabling 100,000-fold faster screening across ultra-large compound libraries. In a prospective screen across ten disease-relevant targets, AuroBind achieved experimental hit rates of 7-69%, with top compounds reaching sub-nanomolar to picomolar potency. For the orphan GPCRs GPR151 and GPR160, AuroBind identified both agonists and antagonists with success rates of 16-30%, and functional assays confirmed GPR160 modulation in liver and prostate cancer models. AuroBind offers a generalizable framework for structure-function learning and high-throughput molecular screening, bridging the gap between structure prediction and therapeutic discovery.

cs.LG↗

arXiv2025

Layer-Aware Representation Filtering: Purifying Finetuning Data to Preserve LLM Safety Alignment

With rapid advancement and increasing accessibility of LLMs, fine-tuning aligned models has become a critical step for adapting them to real-world applications, which makes the safety of this fine-tuning process more important than ever. However, recent studies have highlighted a critical challenge: even when fine-tuning with seemingly benign downstream datasets, the safety of aligned LLMs can be compromised, making them more susceptible to malicious instructions. In this paper, we show that fine-tuning datasets often contain samples with safety-degrading features that are not easily identifiable on the surface. These samples can significantly degrade the safety alignment of LLMs during fine-tuning. To address this issue, we propose LARF, a Layer-Aware Representation Filtering method. This method identifies safety-sensitive layers within the LLM and leverages their representations to detect which data samples in the post-training dataset contain safety-degrading features. Experimental results demonstrate that LARF can effectively identify benign data with safety-degrading features. After removing such data, the safety alignment degradation caused by fine-tuning is mitigated. Please see our code at https://github.com/LLLeoLi/LARF.

cs.CR↗

arXiv2025

The $10~\rm{kpc}$ collar of early-type galaxies -- probing evolution by focusing on the inner stellar density profile

Aims: In this work we aim to investigate the late evolution of ETGs, both observationally and theoretically, by focusing on the stellar mass density profile inside a fixed aperture, within $10 ~\rm{kpc}$ from the galaxy center. Methods: We first studied the stellar mass and mass-weighted density slope within $10 ~\rm{kpc}$, respectively $M_{*,10}$ and $\mathit{Γ_{*,10}}$, of a sample of early-type galaxies from the GAMA survey. We measured the $\mathit{Γ_{*,10}} - M_{*,10}$ relation and its evolution over the redshift range $0.17\leq z \leq 0.37$. We then built a toy model for the merger evolution of galaxies, based on N-body simulations, to explore to what extent the observed growth in $\mathit{Γ_{*,10}} - M_{*,10}$ relation is consistent with a dry-merger evolution scenario. Results: From the observations, we do not detect evidence for an evolution of the $\mathit{Γ_{*,10}} - M_{*,10}$ relation. We put an upper limit on the redshift derivative of the normalization$~(μ)$ and slope$~(β)$of the $\mathit{Γ_{*,10}} - M_{*,10}$ relation $|\partial μ/\partial \log (1+z)| \leq 0.13$ and $\left|\partial β/\partial \log (1+z)\right| \leq 1.10$, respectively. Simulations show that most mergers induce a decrease in $\mathit{Γ_{*,10}}$ and an increase in $M_{*,10}$, although some show a decrease in $M_{*,10}$, particularly for the most extended galaxies and smaller merger mass ratios. By combining the observations with our merger toy model, we placed an upper limit on $f_M = 11.2 \%$ in the redshift range $0.17 \leq z \leq 0.37$. Conclusions: While our measurement is limited by systematics, the application of our approach to samples with a larger redshift baseline, particularly with a time interval $Δt \geq 3.2~\mathrm{Gyr}$, should enable us to detect a signal and help us better understand the late growth of ETGs.

astro-ph.GA↗

arXiv2025

Streamlining Prediction in Bayesian Deep Learning

The rising interest in Bayesian deep learning (BDL) has led to a plethora of methods for estimating the posterior distribution. However, efficient computation of inferences, such as predictions, has been largely overlooked with Monte Carlo integration remaining the standard. In this work we examine streamlining prediction in BDL through a single forward pass without sampling. For this we use local linearisation on activation functions and local Gaussian approximations at linear layers. Thus allowing us to analytically compute an approximation to the posterior predictive distribution. We showcase our approach for both MLP and transformers, such as ViT and GPT-2, and assess its performance on regression and classification tasks. Open-source library: https://github.com/AaltoML/SUQ

cs.LG↗

arXiv2025

How Far are LLMs from Being Our Digital Twins? A Benchmark for Persona-Based Behavior Chain Simulation

Recently, LLMs have garnered increasing attention across academic disciplines for their potential as human digital twins, virtual proxies designed to replicate individuals and autonomously perform tasks such as decision-making, problem-solving, and reasoning on their behalf. However, current evaluations of LLMs primarily emphasize dialogue simulation while overlooking human behavior simulation, which is crucial for digital twins. To address this gap, we introduce BehaviorChain, the first benchmark for evaluating LLMs' ability to simulate continuous human behavior. BehaviorChain comprises diverse, high-quality, persona-based behavior chains, totaling 15,846 distinct behaviors across 1,001 unique personas, each with detailed history and profile metadata. For evaluation, we integrate persona metadata into LLMs and employ them to iteratively infer contextually appropriate behaviors within dynamic scenarios provided by BehaviorChain. Comprehensive evaluation results demonstrated that even state-of-the-art models struggle with accurately simulating continuous human behavior.

cs.CL↗

arXiv2025

Towards Harmonized Uncertainty Estimation for Large Language Models

To facilitate robust and trustworthy deployment of large language models (LLMs), it is essential to quantify the reliability of their generations through uncertainty estimation. While recent efforts have made significant advancements by leveraging the internal logic and linguistic features of LLMs to estimate uncertainty scores, our empirical analysis highlights the pitfalls of these methods to strike a harmonized estimation between indication, balance, and calibration, which hinders their broader capability for accurate uncertainty estimation. To address this challenge, we propose CUE (Corrector for Uncertainty Estimation): A straightforward yet effective method that employs a lightweight model trained on data aligned with the target LLM's performance to adjust uncertainty scores. Comprehensive experiments across diverse models and tasks demonstrate its effectiveness, which achieves consistent improvements of up to 60% over existing methods.

cs.CL↗

arXiv2025

The Giant Radio Array for Neutrino Detection (GRAND) Collaboration -- Contributions to the 39th International Cosmic Ray Conference (ICRC 2025)

The Giant Radio Array for Neutrino Detection (GRAND) is an envisioned observatory of ultra-high-energy particles of cosmic origin, with energies in excess of 100 PeV. GRAND uses large surface arrays of antennas to look for the radio emission from extensive air showers that are triggered by the interaction of ultra-high-energy cosmic rays, gamma rays, and neutrinos in the atmosphere or underground. In particular, for ultra-high-energy neutrinos, the future final phase of GRAND aims to be sensitive enough to detect them in spite of their plausibly tiny flux. Three prototype GRAND radio arrays have been in operation since 2023: GRANDProto300, in China, GRAND@Auger, in Argentina, and GRAND@Nançay, in France. Their goals are to field-test the GRAND detection units, understand the radio background to which they are exposed, and develop tools for diagnostic, data gathering, and data analysis. This list of contributions to the 39th International Cosmic Ray Conference (ICRC 2025) presents an overview of GRAND, in its present and future incarnations, and a first look at data collected by GRANDProto300 and GRAND@Auger, including the first cosmic-ray candidates detected by them.

astro-ph.IM↗

arXiv2025

HotSwap: Enabling Live Dependency Sharing in Serverless Computing

This work presents HotSwap, a novel provider-side cold-start optimization for serverless computing. This optimization reduces cold-start time when booting and loading dependencies at runtime inside a function container. Previous research has extensively focused on reducing cold-start latency for specific functions. However, little attention has been given to skewed production workloads. In such cases, cross-function optimization becomes essential. Without cross-function optimization, a cloud provider is left with two equally poor options: (i) Either the cloud provider gives up optimization for each function in the long tail (which is slow); or (ii) the cloud provider applies function-specific optimizations (e.g., cache function images) to every function in the long tail (which violates the vendor's cache constraints). HotSwap demonstrates cross-function optimization using a novel pre-warming strategy. In this strategy, a pre-initialized live dependency image is migrated to the new function instance. At the same time, HotSwap respects the provider's cache constraints, because a single pre-warmed dependency image in the cache can be shared among all serverless functions that require that image. HotSwap has been tested on seven representative functions from FunctionBench. In those tests, HotSwap accelerates dependency loading for those serverless functions with large dependency requirements by a factor ranging from 2.2 to 3.2. Simulation experiments using Azure traces indicate that HotSwap can save 88\% of space, compared with a previous function-specific method, PreBaking, when sharing a dependency image among ten different functions.

cs.DC↗

arXiv2025

DeFusion: An Effective Decoupling Fusion Network for Multi-Modal Pregnancy Prediction

Temporal embryo images and parental fertility table indicators are both valuable for pregnancy prediction in \textbf{in vitro fertilization embryo transfer} (IVF-ET). However, current machine learning models cannot make full use of the complementary information between the two modalities to improve pregnancy prediction performance. In this paper, we propose a Decoupling Fusion Network called DeFusion to effectively integrate the multi-modal information for IVF-ET pregnancy prediction. Specifically, we propose a decoupling fusion module that decouples the information from the different modalities into related and unrelated information, thereby achieving a more delicate fusion. And we fuse temporal embryo images with a spatial-temporal position encoding, and extract fertility table indicator information with a table transformer. To evaluate the effectiveness of our model, we use a new dataset including 4046 cases collected from Southern Medical University. The experiments show that our model outperforms state-of-the-art methods. Meanwhile, the performance on the eye disease prediction dataset reflects the model's good generalization. Our code is available at https://github.com/Ou-Young-1999/DFNet.

cs.CV↗

arXiv2025

Latent Posterior-Mean Rectified Flow for Higher-Fidelity Perceptual Face Restoration

The Perception-Distortion tradeoff (PD-tradeoff) theory suggests that face restoration algorithms must balance perceptual quality and fidelity. To achieve minimal distortion while maintaining perfect perceptual quality, Posterior-Mean Rectified Flow (PMRF) proposes a flow based approach where source distribution is minimum distortion estimations. Although PMRF is shown to be effective, its pixel-space modeling approach limits its ability to align with human perception, where human perception is defined as how humans distinguish between two image distributions. In this work, we propose Latent-PMRF, which reformulates PMRF in the latent space of a variational autoencoder (VAE), facilitating better alignment with human perception during optimization. By defining the source distribution on latent representations of minimum distortion estimation, we bound the minimum distortion by the VAE's reconstruction error. Moreover, we reveal the design of VAE is crucial, and our proposed VAE significantly outperforms existing VAEs in both reconstruction and restoration. Extensive experiments on blind face restoration demonstrate the superiority of Latent-PMRF, offering an improved PD-tradeoff compared to existing methods, along with remarkable convergence efficiency, achieving a 5.79X speedup over PMRF in terms of FID. Our code will be available as open-source.

cs.CV↗

arXiv2025

2024 NASA SUITS Report: LLM-Driven Immersive Augmented Reality User Interface for Robotics and Space Exploration

As modern computing advances, new interaction paradigms have emerged, particularly in Augmented Reality (AR), which overlays virtual interfaces onto physical objects. This evolution poses challenges in machine perception, especially for tasks like 3D object pose estimation in complex, dynamic environments. Our project addresses critical issues in human-robot interaction within mobile AR, focusing on non-intrusive, spatially aware interfaces. We present URSA, an LLM-driven immersive AR system developed for NASA's 2023-2024 SUITS challenge, targeting future spaceflight needs such as the Artemis missions. URSA integrates three core technologies: a head-mounted AR device (e.g., HoloLens) for intuitive visual feedback, voice control powered by large language models for hands-free interaction, and robot tracking algorithms that enable accurate 3D localization in dynamic settings. To enhance precision, we leverage digital twin localization technologies, using datasets like DTTD-Mobile and specialized hardware such as the ZED2 camera for real-world tracking under noise and occlusion. Our system enables real-time robot control and monitoring via an AR interface, even in the absence of ground-truth sensors--vital for hazardous or remote operations. Key contributions include: (1) a non-intrusive AR interface with LLM-based voice input; (2) a ZED2-based dataset tailored for non-rigid robotic bodies; (3) a Local Mission Control Console (LMCC) for mission visualization; (4) a transformer-based 6DoF pose estimator (DTTDNet) optimized for depth fusion and real-time tracking; and (5) end-to-end integration for astronaut mission support. This work advances digital twin applications in robotics, offering scalable solutions for both aerospace and industrial domains.

cs.RO↗

arXiv2025

Characterizing Many-body Dynamics with Projected Ensembles on a Superconducting Quantum Processor

Quantum simulators offer a new opportunity for the experimental exploration of non-equilibrium quantum many-body dynamics, which have traditionally been characterized through expectation values or entanglement measures, based on density matrices of the system. Recently, a more general framework for studying quantum many-body systems based on projected ensembles has been introduced, revealing novel quantum phenomena, such as deep thermalization in chaotic systems. Here, we experimentally investigate a chaotic quantum many-body system using projected ensembles on a superconducting processor with 16 qubits on a square lattice. Our results provide direct evidence of deep thermalization by observing a Haar-distributed projected ensemble for the steady states within a charge-conserved sector. Moreover, by introducing an ensemble-averaged entropy as a metric, we establish a benchmark for many-body information leakage from the system to its environment. Our work paves the way for studying quantum many-body dynamics using projected ensembles and shows a potential implication for advancing quantum simulation techniques.

quant-ph↗

arXiv2025

Position: Intelligent Science Laboratory Requires the Integration of Cognitive and Embodied AI

Scientific discovery has long been constrained by human limitations in expertise, physical capability, and sleep cycles. The recent rise of AI scientists and automated laboratories has accelerated both the cognitive and operational aspects of research. However, key limitations persist: AI systems are often confined to virtual environments, while automated laboratories lack the flexibility and autonomy to adaptively test new hypotheses in the physical world. Recent advances in embodied AI, such as generalist robot foundation models, diffusion-based action policies, fine-grained manipulation learning, and sim-to-real transfer, highlight the promise of integrating cognitive and embodied intelligence. This convergence opens the door to closed-loop systems that support iterative, autonomous experimentation and the possibility of serendipitous discovery. In this position paper, we propose the paradigm of Intelligent Science Laboratories (ISLs): a multi-layered, closed-loop framework that deeply integrates cognitive and embodied intelligence. ISLs unify foundation models for scientific reasoning, agent-based workflow orchestration, and embodied agents for robust physical experimentation. We argue that such systems are essential for overcoming the current limitations of scientific discovery and for realizing the full transformative potential of AI-driven science.

cs.AI↗

arXiv2025

IsoNet: Causal Analysis of Multimodal Transformers for Neuromuscular Gesture Classification

Hand gestures are a primary output of the human motor system, yet the decoding of their neuromuscular signatures remains a bottleneck for basic neuroscience and assistive technologies such as prosthetics. Traditional human-machine interface pipelines rely on a single biosignal modality, but multimodal fusion can exploit complementary information from sensors. We systematically compare linear and attention-based fusion strategies across three architectures: a Multimodal MLP, a Multimodal Transformer, and a Hierarchical Transformer, evaluating performance on scenarios with unimodal and multimodal inputs. Experiments use two publicly available datasets: NinaPro DB2 (sEMG and accelerometer) and HD-sEMG 65-Gesture (high-density sEMG and force). Across both datasets, the Hierarchical Transformer with attention-based fusion consistently achieved the highest accuracy, surpassing the multimodal and best single-modality linear-fusion MLP baseline by over 10% on NinaPro DB2 and 3.7% on HD-sEMG. To investigate how modalities interact, we introduce an Isolation Network that selectively silences unimodal or cross-modal attention pathways, quantifying each group of token interactions' contribution to downstream decisions. Ablations reveal that cross-modal interactions contribute approximately 30% of the decision signal across transformer layers, highlighting the importance of attention-driven fusion in harnessing complementary modality information. Together, these findings reveal when and how multimodal fusion would enhance biosignal classification and also provides mechanistic insights of human muscle activities. The study would be beneficial in the design of sensor arrays for neurorobotic systems.

cs.LG↗

arXiv2025

A Language-Agnostic Logical Relation for Message-Passing Protocols

Today's computing landscape has been gradually shifting to applications targeting distributed and *heterogeneous* systems, such as cloud computing and Internet of Things (IoT) applications. These applications are predominantly *concurrent*, employ *message-passing*, and interface with *foreign objects*, ranging from externally implemented code to actual physical devices such as sensors. Verifying that the resulting systems adhere to the intended protocol of interaction is challenging -- the usual assumption of a common implementation language, let alone a type system, no longer applies, ruling out any verification method based on them. This paper develops a framework for certifying *protocol compliance* of heterogeneous message-passing systems. It contributes the first mechanization of a *language-agnostic logical relation*, asserting that its inhabitants comply with the protocol specified. This definition relies entirely on a labelled transition-based semantics, accommodating arbitrary inhabitants, typed and untyped alike, including foreign objects. As a case study, the paper considers two scenarios: (1) *per-instance verification* of a specific application or hardware device, and (2) *once-and-for-all verification* of well-typed applications for a given type system. The logical relation and both scenarios are mechanized in the Coq theorem prover.

cs.PL↗

arXiv2025

Denoising Programming Knowledge Tracing with a Code Graph-based Tuning Adaptor

Programming Knowledge Tracking (PKT) aims to dynamically diagnose learners' mastery levels of programming knowledge based on their coding activities, facilitating more effective and personalized programming education. However, current PKT studies primarily focus on the implicit relationship between code content and knowledge assessment, often overlooking two types of noise signals in long-term programming activities: unwanted signals from unrelated submissions and weak signals from minor modifications. This practical challenge significantly limits model performance and application. To address this issue, we propose Coda, a Code graph-based tuning adaptor designed to enhance existing PKT models by identifying and mitigating the impact of noise. Specifically, Coda first transforms the loose code sequences submitted by each learner into a compact code graph. By leveraging this code graph, unwanted signals can be identified from a semantic similarity perspective. We then apply a cluster-aware GCN to the code graph, which improves the discrimination of weak signals and enables their clustering for identification. Finally, a lightweight yet effective adaptor is incorporated into the PKT task through optimization with two noise feature-based constraints and a navigational regularization term, to correct knowledge states affected by noise. It is worth mentioning that the Coda framework is model-agnostic and can be adapted to most existing PKT solutions. Extensive experimental results on four real-world datasets demonstrate that Coda effectively performs the PKT task in the presence of noisy programming records, outperforming typical baselines.

cs.SE↗