Search arXiv⌕ Search

arXiv subjects

Yi Liu

Publications and source records attributed to Yi Liu.

At least 487 records · Page 27Linked to original sources
arXiv2024

Pandora: Jailbreak GPTs by Retrieval Augmented Generation Poisoning

Large Language Models~(LLMs) have gained immense popularity and are being increasingly applied in various domains. Consequently, ensuring the security of these models is of paramount importance. Jailbreak attacks, which manipulate LLMs to generate malicious content, are recognized as a significant vulnerability. While existing research has predominantly focused on direct jailbreak attacks on LLMs, there has been limited exploration of indirect methods. The integration of various plugins into LLMs, notably Retrieval Augmented Generation~(RAG), which enables LLMs to incorporate external knowledge bases into their response generation such as GPTs, introduces new avenues for indirect jailbreak attacks. To fill this gap, we investigate indirect jailbreak attacks on LLMs, particularly GPTs, introducing a novel attack vector named Retrieval Augmented Generation Poisoning. This method, Pandora, exploits the synergy between LLMs and RAG through prompt manipulation to generate unexpected responses. Pandora uses maliciously crafted content to influence the RAG process, effectively initiating jailbreak attacks. Our preliminary tests show that Pandora successfully conducts jailbreak attacks in four different scenarios, achieving higher success rates than direct attacks, with 64.3\% for GPT-3.5 and 34.8\% for GPT-4.

cs.CR↗
arXiv2024

Efficient Preference-Based Reinforcement Learning Using Learned Dynamics Models

Preference-based reinforcement learning (PbRL) can enable robots to learn to perform tasks based on an individual's preferences without requiring a hand-crafted reward function. However, existing approaches either assume access to a high-fidelity simulator or analytic model or take a model-free approach that requires extensive, possibly unsafe online environment interactions. In this paper, we study the benefits and challenges of using a learned dynamics model when performing PbRL. In particular, we provide evidence that a learned dynamics model offers the following benefits when performing PbRL: (1) preference elicitation and policy optimization require significantly fewer environment interactions than model-free PbRL, (2) diverse preference queries can be synthesized safely and efficiently as a byproduct of standard model-based RL, and (3) reward pre-training based on suboptimal demonstrations can be performed without any environmental interaction. Our paper provides empirical evidence that learned dynamics models enable robots to learn customized policies based on user preferences in ways that are safer and more sample efficient than prior preference learning approaches. Supplementary materials and code are available at https://sites.google.com/berkeley.edu/mop-rl.

cs.LG↗
arXiv2024

Distinct pressure evolution of superconductivity and charge-density-wave in kagome superconductor CsV$_3$Sb$_5$ thin flakes

It is intriguing to explore the coexistence and (or) competition between charge-density-wave (CDW) and superconductivity (SC) in many correlated electron systems, such as cuprates, organic superconductors and dichacolgenides. Among them, the recently discovered $\mathbb{Z} _2$ topological kagome metals AV$_3$Sb$_5$ (A=K, Rb, Cs) serve as an ideal platform to study the intricate relation between them. Here, we report the electrical resistance measurements on CsV$_3$Sb$_5$ thin flakes ($\approx$ 60 nm) under hydrostatic pressure up to 2.12 GPa to compare its pressure phase diagram of CDW and SC with its bulk form. Even though the CDW transition temperature (T$_{CDW}$) in CsV$_3$Sb$_5$ thin flakes is still monotonically suppressed under pressure and totally vanishes at P$_2$=1.83 GPa similar to the bulk, the superconducting transition temperature (T$_c$) shows an initial decrease and consequent increase up to its maximum $\sim$ 8.03 K at P$_2$, in sharp contrast with the M-shaped double domes in the bulk CsV$_3$Sb$_5$. Our results suggest the important role of reduced dimensionality on the CDW state and its interplay with the SC, offering a new perspective to explore the exotic nature of CsV$_3$Sb$_5$.

cond-mat.supr-con↗
arXiv2024

Overlap, matching, or entropy weights: what are we weighting for?

There has been a recent surge in statistical methods for handling the lack of adequate positivity when using inverse probability weights (IPW). However, these nascent developments have raised a number of questions. Thus, we demonstrate the ability of equipoise estimators (overlap, matching, and entropy weights) to handle the lack of positivity. Compared to IPW, the equipoise estimators have been shown to be flexible and easy to interpret. However, promoting their wide use requires that researchers know clearly why, when to apply them and what to expect. In this paper, we provide the rationale to use these estimators to achieve robust results. We specifically look into the impact imbalances in treatment allocation can have on the positivity and, ultimately, on the estimates of the treatment effect. We zero into the typical pitfalls of the IPW estimator and its relationship with the estimators of the average treatment effect on the treated (ATT) and on the controls (ATC). Furthermore, we also compare IPW trimming to the equipoise estimators. We focus particularly on two key points: What fundamentally distinguishes their estimands? When should we expect similar results? Our findings are illustrated through Monte-Carlo simulation studies and a data example on healthcare expenditure.

stat.ME↗
arXiv2024

On Hempel pairs and Turaev--Viro invariants

Surface bundles arising from periodic mapping classes may sometimes have non-isomorphic, but profinitely isomorphic fundamental groups. Pairs of this kind have been discovered by Hempel. This paper exhibits examples of nontrivial Hempel pairs where the mapping tori can be distinguished by some Turaev--Viro invariants, and also examples where they cannot be distinguished by any Turaev--Viro invariants.

math.GT↗
arXiv2024

Efficiently Solving High-Order and Nonlinear ODEs with Rational Fraction Polynomial: the Ratio Net

Recent advances in solving ordinary differential equations (ODEs) with neural networks have been remarkable. Neural networks excel at serving as trial functions and approximating solutions within functional spaces, aided by gradient backpropagation algorithms. However, challenges remain in solving complex ODEs, including high-order and nonlinear cases, emphasizing the need for improved efficiency and effectiveness. Traditional methods have typically relied on established knowledge integration to improve problem-solving efficiency. In contrast, this study takes a different approach by introducing a new neural network architecture for constructing trial functions, known as ratio net. This architecture draws inspiration from rational fraction polynomial approximation functions, specifically the Pade approximant. Through empirical trials, it demonstrated that the proposed method exhibits higher efficiency compared to existing approaches, including polynomial-based and multilayer perceptron (MLP) neural network-based methods. The ratio net holds promise for advancing the efficiency and effectiveness of solving differential equations.

cs.LG↗
arXiv2024

A Cross-Language Investigation into Jailbreak Attacks in Large Language Models

Large Language Models (LLMs) have become increasingly popular for their advanced text generation capabilities across various domains. However, like any software, they face security challenges, including the risk of 'jailbreak' attacks that manipulate LLMs to produce prohibited content. A particularly underexplored area is the Multilingual Jailbreak attack, where malicious questions are translated into various languages to evade safety filters. Currently, there is a lack of comprehensive empirical studies addressing this specific threat. To address this research gap, we conducted an extensive empirical study on Multilingual Jailbreak attacks. We developed a novel semantic-preserving algorithm to create a multilingual jailbreak dataset and conducted an exhaustive evaluation on both widely-used open-source and commercial LLMs, including GPT-4 and LLaMa. Additionally, we performed interpretability analysis to uncover patterns in Multilingual Jailbreak attacks and implemented a fine-tuning mitigation method. Our findings reveal that our mitigation strategy significantly enhances model defense, reducing the attack success rate by 96.2%. This study provides valuable insights into understanding and mitigating Multilingual Jailbreak attacks.

cs.CR↗
arXiv2024

Digger: Detecting Copyright Content Mis-usage in Large Language Model Training

Pre-training, which utilizes extensive and varied datasets, is a critical factor in the success of Large Language Models (LLMs) across numerous applications. However, the detailed makeup of these datasets is often not disclosed, leading to concerns about data security and potential misuse. This is particularly relevant when copyrighted material, still under legal protection, is used inappropriately, either intentionally or unintentionally, infringing on the rights of the authors. In this paper, we introduce a detailed framework designed to detect and assess the presence of content from potentially copyrighted books within the training datasets of LLMs. This framework also provides a confidence estimation for the likelihood of each content sample's inclusion. To validate our approach, we conduct a series of simulated experiments, the results of which affirm the framework's effectiveness in identifying and addressing instances of content misuse in LLM training processes. Furthermore, we investigate the presence of recognizable quotes from famous literary works within these datasets. The outcomes of our study have significant implications for ensuring the ethical use of copyrighted materials in the development of LLMs, highlighting the need for more transparent and responsible data management practices in this field.

cs.CR↗
arXiv2024

Benchmarking Large Language Models on Controllable Generation under Diversified Instructions

While large language models (LLMs) have exhibited impressive instruction-following capabilities, it is still unclear whether and to what extent they can respond to explicit constraints that might be entailed in various instructions. As a significant aspect of LLM alignment, it is thus important to formulate such a specialized set of instructions as well as investigate the resulting behavior of LLMs. To address this vacancy, we propose a new benchmark CoDI-Eval to systematically and comprehensively evaluate LLMs' responses to instructions with various constraints. We construct a large collection of constraints-attributed instructions as a test suite focused on both generalization and coverage. Specifically, we advocate an instruction diversification process to synthesize diverse forms of constraint expression and also deliberate the candidate task taxonomy with even finer-grained sub-categories. Finally, we automate the entire evaluation process to facilitate further developments. Different from existing studies on controllable text generation, CoDI-Eval extends the scope to the prevalent instruction-following paradigm for the first time. We provide extensive evaluations of representative LLMs (e.g., ChatGPT, Vicuna) on CoDI-Eval, revealing their limitations in following instructions with specific constraints and there is still a significant gap between open-source and commercial closed-source LLMs. We believe this benchmark will facilitate research into improving the controllability of LLMs' responses to instructions. Our data and code are available at https://github.com/Xt-cyh/CoDI-Eval.

cs.CL↗
arXiv2024

Towards Efficient and Effective Text-to-Video Retrieval with Coarse-to-Fine Visual Representation Learning

In recent years, text-to-video retrieval methods based on CLIP have experienced rapid development. The primary direction of evolution is to exploit the much wider gamut of visual and textual cues to achieve alignment. Concretely, those methods with impressive performance often design a heavy fusion block for sentence (words)-video (frames) interaction, regardless of the prohibitive computation complexity. Nevertheless, these approaches are not optimal in terms of feature utilization and retrieval efficiency. To address this issue, we adopt multi-granularity visual feature learning, ensuring the model's comprehensiveness in capturing visual content features spanning from abstract to detailed levels during the training phase. To better leverage the multi-granularity features, we devise a two-stage retrieval architecture in the retrieval phase. This solution ingeniously balances the coarse and fine granularity of retrieval content. Moreover, it also strikes a harmonious equilibrium between retrieval effectiveness and efficiency. Specifically, in training phase, we design a parameter-free text-gated interaction block (TIB) for fine-grained video representation learning and embed an extra Pearson Constraint to optimize cross-modal representation learning. In retrieval phase, we use coarse-grained video representations for fast recall of top-k candidates, which are then reranked by fine-grained video representations. Extensive experiments on four benchmarks demonstrate the efficiency and effectiveness. Notably, our method achieves comparable performance with the current state-of-the-art methods while being nearly 50 times faster.

cs.CV↗
arXiv2023

Quantum phase transitions in two-dimensional superconductors: a review on recent experimental progress

Superconductor-insulator/metal transition (SIT/SMT) as a paradigm of quantum phase transition has been a research highlight over the last three decades. Benefit from recent developments in the fabrication and measurements of 2D superconducting films and nanodevices, unprecedented quantum phenomena have been revealed in the quantum phase transitions of 2D superconductors. In this review, we introduce the recent progress on quantum phase transitions in 2D superconductors, focusing on the quantum Griffiths singularity (QGS) and anomalous metal state. Characterized by a divergent critical exponent when approaching zero temperature, QGS of SMT is discovered in ultrathin crystalline Ga films and subsequently detected in various 2D superconductors. The universality of QGS indicates the profound influence of quenched disorder on quantum phase transitions. Besides, in a 2D superconducting system, whether a metallic ground state can exist is a long-sought mystery. Recently, the charge-2e quantum oscillations are observed in nanopatterned superconducting films, indicating the bosonic nature of the anomalous metal state and ending the debate on whether bosons can exist as a metal. The evidences of the anomalous metal states have also been reported in crystalline epitaxial thin films and exfoliated nanoflakes, as well as granular composite films. High quality filters are used in these works to exclude the influence of external high frequency noises in ultralow temperature measurements. The observations of QGS and metallic ground states in 2D superconductors not only reveal the prominent role of quantum fluctuations and dissipations but also provide new perspective to explore quantum phase transitions in superconducting systems.

cond-mat.supr-con↗
arXiv2023

Superconductivity and Charge-density-wave-like Transition in Th2Cu4As5

We report the synthesis, crystal structure, and physical properties of a novel ternary compound, Th$_2$Cu$_4$As$_5$. The material crystallizes in a tetragonal structure with lattice parameters $a=4.0716(1)$ Å and $c=24.8131(4)$ Å. Its structure can be described as an alternating stacking of fluorite-type Th$_2$As$_2$ layers with antifluorite-type double-layered Cu$_4$As$_3$ slabs. The measurement of electrical resistivity, magnetic susceptibility and specific heat reveals that Th$_2$Cu$_4$As$_5$ undergoes bulk superconducting transition at 4.2 K. Moreover, all these physical quantities exhibit anomalies at 48 K, where the Hall coefficient change the sign. These findings suggest a charge-density-wave-like (CDW) transition, making Th$_2$Cu$_4$As$_5$ a rare example for studying the interplay between CDW and superconductivity.

cond-mat.supr-con↗
arXiv2023

RECALL: A Benchmark for LLMs Robustness against External Counterfactual Knowledge

LLMs and AI chatbots have improved people's efficiency in various fields. However, the necessary knowledge for answering the question may be beyond the models' knowledge boundaries. To mitigate this issue, many researchers try to introduce external knowledge, such as knowledge graphs and Internet contents, into LLMs for up-to-date information. However, the external information from the Internet may include counterfactual information that will confuse the model and lead to an incorrect response. Thus there is a pressing need for LLMs to possess the ability to distinguish reliable information from external knowledge. Therefore, to evaluate the ability of LLMs to discern the reliability of external knowledge, we create a benchmark from existing knowledge bases. Our benchmark consists of two tasks, Question Answering and Text Generation, and for each task, we provide models with a context containing counterfactual information. Evaluation results show that existing LLMs are susceptible to interference from unreliable external knowledge with counterfactual information, and simple intervention methods make limited contributions to the alleviation of this issue.

cs.CL↗
arXiv2023

Gaussian Differential Privacy on Riemannian Manifolds

We develop an advanced approach for extending Gaussian Differential Privacy (GDP) to general Riemannian manifolds. The concept of GDP stands out as a prominent privacy definition that strongly warrants extension to manifold settings, due to its central limit properties. By harnessing the power of the renowned Bishop-Gromov theorem in geometric analysis, we propose a Riemannian Gaussian distribution that integrates the Riemannian distance, allowing us to achieve GDP in Riemannian manifolds with bounded Ricci curvature. To the best of our knowledge, this work marks the first instance of extending the GDP framework to accommodate general Riemannian manifolds, encompassing curved spaces, and circumventing the reliance on tangent space summaries. We provide a simple algorithm to evaluate the privacy budget $μ$ on any one-dimensional manifold and introduce a versatile Markov Chain Monte Carlo (MCMC)-based algorithm to calculate $μ$ on any Riemannian manifold with constant curvature. Through simulations on one of the most prevalent manifolds in statistics, the unit sphere $S^d$, we demonstrate the superior utility of our Riemannian Gaussian mechanism in comparison to the previously proposed Riemannian Laplace mechanism for implementing GDP.

cs.CR↗
arXiv2023

EasyScale: Accuracy-consistent Elastic Training for Deep Learning

Distributed synchronized GPU training is commonly used for deep learning. The resource constraint of using a fixed number of GPUs makes large-scale training jobs suffer from long queuing time for resource allocation, and lowers the cluster utilization. Adapting to resource elasticity can alleviate this but often introduces inconsistent model accuracy, due to lacking of capability to decouple model training procedure from resource allocation. We propose EasyScale, an elastic training system that achieves consistent model accuracy under resource elasticity for both homogeneous and heterogeneous GPUs. EasyScale preserves the data-parallel training behaviors strictly, traces the consistency-relevant factors carefully, utilizes the deep learning characteristics for EasyScaleThread abstraction and fast context-switching. To utilize heterogeneous cluster, EasyScale dynamically assigns workers based on the intra-/inter-job schedulers, minimizing load imbalance and maximizing aggregated job throughput. Deployed in an online serving cluster, EasyScale powers the training jobs to utilize idle GPUs opportunistically, improving overall cluster utilization by 62.1%.

cs.DC↗
arXiv2023

Efficient Approximations of Complete Interatomic Potentials for Crystal Property Prediction

We study property prediction for crystal materials. A crystal structure consists of a minimal unit cell that is repeated infinitely in 3D space. How to accurately represent such repetitive structures in machine learning models remains unresolved. Current methods construct graphs by establishing edges only between nearby nodes, thereby failing to faithfully capture infinite repeating patterns and distant interatomic interactions. In this work, we propose several innovations to overcome these limitations. First, we propose to model physics-principled interatomic potentials directly instead of only using distances as in many existing methods. These potentials include the Coulomb potential, London dispersion potential, and Pauli repulsion potential. Second, we model the complete set of potentials among all atoms, instead of only between nearby atoms as in existing methods. This is enabled by our approximations of infinite potential summations, where we extend the Ewald summation for several potential series approximations with provable error bounds. Finally, we propose to incorporate our computations of complete interatomic potentials into message passing neural networks for representation learning. We perform experiments on the JARVIS and Materials Project benchmarks for evaluation. Results show that the use of interatomic potentials and complete interatomic potentials leads to consistent performance improvements with reasonable computational costs. Our code is publicly available as part of the AIRS library (https://github.com/divelab/AIRS/tree/main/OpenMat/PotNet).

physics.chem-ph↗
arXiv2023

Strain-Tunable Magnetic Compensation Temperature of Epitaxial Tb$_3$Fe$_5$O$_{12}$ Thin Films

High-quality rare-earth iron garnet (ReIG) Tb$_3$Fe$_5$O$_{12}$ (TbIG) thin films are epitaxially grown on a series of (111)-oriented garnet substrates with various lattice constants. The coherent growth induces a substrate-dependent in-plane tensile or compressive strain in the TbIG film. Measurements of the anomalous Hall-like effect (AHLE) in TbIG/Pt heterostructures show that the compensation temperature of TbIG films monotonically changes with the film strain. The strain results in a variation of the distances between magnetic atoms in the TbIG crystal and therefore the corresponding exchange interactions. The latter is explicitly calculated as a function of the lattice strain based on density functional theory, reproducing the observed experimental results. This work provides a versatile way to optimize ReIG-based spin-orbit torque devices.

cond-mat.mtrl-sci↗
arXiv2023

MasterKey: Automated Jailbreak Across Multiple Large Language Model Chatbots

Large Language Models (LLMs) have revolutionized Artificial Intelligence (AI) services due to their exceptional proficiency in understanding and generating human-like text. LLM chatbots, in particular, have seen widespread adoption, transforming human-machine interactions. However, these LLM chatbots are susceptible to "jailbreak" attacks, where malicious users manipulate prompts to elicit inappropriate or sensitive responses, contravening service policies. Despite existing attempts to mitigate such threats, our research reveals a substantial gap in our understanding of these vulnerabilities, largely due to the undisclosed defensive measures implemented by LLM service providers. In this paper, we present Jailbreaker, a comprehensive framework that offers an in-depth understanding of jailbreak attacks and countermeasures. Our work makes a dual contribution. First, we propose an innovative methodology inspired by time-based SQL injection techniques to reverse-engineer the defensive strategies of prominent LLM chatbots, such as ChatGPT, Bard, and Bing Chat. This time-sensitive approach uncovers intricate details about these services' defenses, facilitating a proof-of-concept attack that successfully bypasses their mechanisms. Second, we introduce an automatic generation method for jailbreak prompts. Leveraging a fine-tuned LLM, we validate the potential of automated jailbreak generation across various commercial LLM chatbots. Our method achieves a promising average success rate of 21.58%, significantly outperforming the effectiveness of existing techniques. We have responsibly disclosed our findings to the concerned service providers, underscoring the urgent need for more robust defenses. Jailbreaker thus marks a significant step towards understanding and mitigating jailbreak threats in the realm of LLM chatbots.

cs.CR↗