arXiv subjects

Kaixuan Huang

Publications and source records attributed to Kaixuan Huang.

At least 55 records · Page 3Linked to original sources

arXiv2025

MATH-Perturb: Benchmarking LLMs' Math Reasoning Abilities against Hard Perturbations

Large language models have demonstrated impressive performance on challenging mathematical reasoning tasks, which has triggered the discussion of whether the performance is achieved by true reasoning capability or memorization. To investigate this question, prior work has constructed mathematical benchmarks when questions undergo simple perturbations -- modifications that still preserve the underlying reasoning patterns of the solutions. However, no work has explored hard perturbations, which fundamentally change the nature of the problem so that the original solution steps do not apply. To bridge the gap, we construct MATH-P-Simple and MATH-P-Hard via simple perturbation and hard perturbation, respectively. Each consists of 279 perturbed math problems derived from level-5 (hardest) problems in the MATH dataset (Hendrycksmath et. al., 2021). We observe significant performance drops on MATH-P-Hard across various models, including o1-mini (-16.49%) and gemini-2.0-flash-thinking (-12.9%). We also raise concerns about a novel form of memorization where models blindly apply learned problem-solving skills without assessing their applicability to modified contexts. This issue is amplified when using original problems for in-context learning. We call for research efforts to address this challenge, which is critical for developing more robust and reliable reasoning models.

cs.LG↗

arXiv2024

A Theoretical Perspective for Speculative Decoding Algorithm

Transformer-based autoregressive sampling has been the major bottleneck for slowing down large language model inferences. One effective way to accelerate inference is \emph{Speculative Decoding}, which employs a small model to sample a sequence of draft tokens and a large model to validate. Given its empirical effectiveness, the theoretical understanding of Speculative Decoding is falling behind. This paper tackles this gap by conceptualizing the decoding problem via markov chain abstraction and studying the key properties, \emph{output quality and inference acceleration}, from a theoretical perspective. Our analysis covers the theoretical limits of speculative decoding, batch algorithms, and output quality-inference acceleration tradeoffs. Our results reveal the fundamental connections between different components of LLMs via total variation distances and show how they jointly affect the efficiency of decoding algorithms.

cs.LG↗

arXiv2024

Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications

Large language models (LLMs) show inherent brittleness in their safety mechanisms, as evidenced by their susceptibility to jailbreaking and even non-malicious fine-tuning. This study explores this brittleness of safety alignment by leveraging pruning and low-rank modifications. We develop methods to identify critical regions that are vital for safety guardrails, and that are disentangled from utility-relevant regions at both the neuron and rank levels. Surprisingly, the isolated regions we find are sparse, comprising about $3\%$ at the parameter level and $2.5\%$ at the rank level. Removing these regions compromises safety without significantly impacting utility, corroborating the inherent brittleness of the model's safety mechanisms. Moreover, we show that LLMs remain vulnerable to low-cost fine-tuning attacks even when modifications to the safety-critical regions are restricted. These findings underscore the urgent need for more robust safety strategies in LLMs.

cs.LG↗

arXiv2024

High-precision pulse calibration of tunable couplers for high-fidelity two-qubit gates in superconducting quantum processors

For superconducting quantum processors, stable high-fidelity two-qubit operations depend on precise flux control of the tunable coupler. However, the pulse distortion poses a significant challenge to the control precision. Current calibration methods, which often rely on microwave crosstalk or additional readout resonators for coupler excitation and readout, tend to be cumbersome and inefficient, especially when couplers only have flux control. Here, we introduce and experimentally validate a novel pulse calibration scheme that exploits the strong coupling between qubits and couplers, eliminating the need for extra coupler readout and excitation. Our method directly measures the short-time and long-time step responses of the coupler flux pulse transient, enabling us to apply predistortion to subsequent signals using fast Fourier transformation and deconvolution. This approach not only simplifies the calibration process but also significantly improves the precision and stability of the flux control. We demonstrate the efficacy of our method through the implementation of diabatic CZ and iSWAP gates with fidelities of $99.61\pm0.04\%$ and $99.82\pm0.02\%$, respectively, as well as a series of diabatic CPhase gates with high fidelities characterized by cross-entropy benchmarking. The consistency and robustness of our technique are further validated by the reduction in pulse distortion and phase error observed across multilayer CZ gates. These results underscore the potential of our calibration and predistortion method to enhance the performance of two-qubit gates in superconducting quantum processors.

quant-ph↗

arXiv2024

Latent Diffusion Models for Controllable RNA Sequence Generation

This work presents RNAdiffusion, a latent diffusion model for generating and optimizing discrete RNA sequences of variable lengths. RNA is a key intermediary between DNA and protein, exhibiting high sequence diversity and complex three-dimensional structures to support a wide range of functions. We utilize pretrained BERT-type models to encode raw RNA sequences into token-level, biologically meaningful representations. A Query Transformer is employed to compress such representations into a set of fixed-length latent vectors, with an autoregressive decoder trained to reconstruct RNA sequences from these latent variables. We then develop a continuous diffusion model within this latent space. To enable optimization, we integrate the gradients of reward models--surrogates for RNA functional properties--into the backward diffusion process, thereby generating RNAs with high reward scores. Empirical results confirm that RNAdiffusion generates non-coding RNAs that align with natural distributions across various biological metrics. Further, we fine-tune the diffusion model on mRNA 5' untranslated regions (5'-UTRs) and optimize sequences for high translation efficiencies. Our guided diffusion model effectively generates diverse 5'-UTRs with high Mean Ribosome Loading (MRL) and Translation Efficiency (TE), outperforming baselines in balancing rewards and structural stability trade-off. Our findings hold potential for advancing RNA sequence-function research and therapeutic RNA design.

cs.LG↗

arXiv2024

Investigations on the equation of state of neutron star matter with density-dependent relativistic mean-field model

The compact object with a mass of $2.50-2.67~M_\odot$ observed by LIGO Scientific and Virgo collaborations in GW190814, as well as the recent report of a light compact object with a mass and radius of $M=0.77^{+0.20}_{-0.17}M_{\odot}$ and $R=10.4^{+0.86}_{-0.78}$ km within the supernova remnant HESS J1731-347, have posed a great challenge to the investigations into the supranuclear matter. In the inner core region of the neutron star, the strangeness degrees of freedom, such as the hyperons, can be present, which is also named as a hyperonic star. In this work, the neutron star consisting of nucleons and leptons, and the hyperonic star including the hyperons will be studied in the framework of the density-dependent relativistic mean-field (DDRMF) model. Some popular DDRMF parameterizations will be adopted to investigate the properties of nuclear matter and the mass, radius, tidal deformability, and other properties of neutron star and hyperonic stars. We find that the maximum masses of neutron star calculated by DD-MEX, DD-MEX1, DD-MEX2, DD-MEXY and DD-LZ1 sets can be around $2.5-2.6~M_\odot$ with quite stiff equations of state (EOSs) generated by their strong repulsive contributions from vector potentials at high densities. Moreover, by investigating the influence of the crust EOS and core EOS on the neutron stars, we find that the observational data from HESS J1731-347 suggest the requirement of a crust EOS with a higher $L$ parameter and a core EOS with a lower $L$ parameter, and the $M-R$ relations from the constructed EOSs can also be consistent with the observables of PSR J0740+6620, PSR J0030+0451 from NICER and the GW170817 event. With the inclusion of hyperons, the hyperonic star matter becomes softer compared to the neutron star matter. But the massive hyperonic star can also be obtained with DDRMF parameter sets if the vector coupling constants are strong.

astro-ph.HE↗

arXiv2024

Probing spin hydrodynamics on a superconducting quantum simulator

Characterizing the nature of hydrodynamical transport properties in quantum dynamics provides valuable insights into the fundamental understanding of exotic non-equilibrium phases of matter. Experimentally simulating infinite-temperature transport on large-scale complex quantum systems is of considerable interest. Here, using a controllable and coherent superconducting quantum simulator, we experimentally realize the analog quantum circuit, which can efficiently prepare the Haar-random states, and probe spin transport at infinite temperature. We observe diffusive spin transport during the unitary evolution of the ladder-type quantum simulator with ergodic dynamics. Moreover, we explore the transport properties of the systems subjected to strong disorder or a tilted potential, revealing signatures of anomalous subdiffusion in accompany with the breakdown of thermalization. Our work demonstrates a scalable method of probing infinite-temperature spin transport on analog quantum simulators, which paves the way to study other intriguing out-of-equilibrium phenomena from the perspective of transport.

quant-ph↗

arXiv2024

Deep Reinforcement Learning for Efficient and Fair Allocation of Health Care Resources

Scarcity of health care resources could result in the unavoidable consequence of rationing. For example, ventilators are often limited in supply, especially during public health emergencies or in resource-constrained health care settings, such as amid the pandemic of COVID-19. Currently, there is no universally accepted standard for health care resource allocation protocols, resulting in different governments prioritizing patients based on various criteria and heuristic-based protocols. In this study, we investigate the use of reinforcement learning for critical care resource allocation policy optimization to fairly and effectively ration resources. We propose a transformer-based deep Q-network to integrate the disease progression of individual patients and the interaction effects among patients during the critical care resource allocation. We aim to improve both fairness of allocation and overall patient outcomes. Our experiments demonstrate that our method significantly reduces excess deaths and achieves a more equitable distribution under different levels of ventilator shortage, when compared to existing severity-based and comorbidity-based methods in use by different governments. Our source code is included in the supplement and will be released on Github upon publication.

cs.LG↗

arXiv2024

Quantum Compiling with Reinforcement Learning on a Superconducting Processor

To effectively implement quantum algorithms on noisy intermediate-scale quantum (NISQ) processors is a central task in modern quantum technology. NISQ processors feature tens to a few hundreds of noisy qubits with limited coherence times and gate operations with errors, so NISQ algorithms naturally require employing circuits of short lengths via quantum compilation. Here, we develop a reinforcement learning (RL)-based quantum compiler for a superconducting processor and demonstrate its capability of discovering novel and hardware-amenable circuits with short lengths. We show that for the three-qubit quantum Fourier transformation, a compiled circuit using only seven CZ gates with unity circuit fidelity can be achieved. The compiler is also able to find optimal circuits under device topological constraints, with lengths considerably shorter than those by the conventional method. Our study exemplifies the codesign of the software with hardware for efficient quantum compilation, offering valuable insights for the advancement of RL-based compilers.

quant-ph↗

arXiv2024

Method for detector description conversion from DD4hep to Filmbox

DD4hep serves as a generic detector description toolkit recommended for offline software development in next-generation high-energy physics~(HEP) experiments. Conversely, Filmbox~(FBX) stands out as a widely used 3D modeling file format within the 3D software industry. In this paper, we introduce a novel method that can automatically convert complex HEP detector geometries from DD4hep description into 3D models in the FBX format. The feasibility of this method was demonstrated by its application to the DD4hep description of the Compact Linear Collider detector and several sub-detectors of the super Tau-Charm facility and circular electron-positron collider experiments. The automatic DD4hep--FBX detector conversion interface provides convenience for further development of applications, such as detector design, simulation, visualization, data monitoring, and outreach, in HEP experiments.

physics.ins-det↗

arXiv2024

AI Risk Management Should Incorporate Both Safety and Security

The exposure of security vulnerabilities in safety-aligned language models, e.g., susceptibility to adversarial attacks, has shed light on the intricate interplay between AI safety and AI security. Although the two disciplines now come together under the overarching goal of AI risk management, they have historically evolved separately, giving rise to differing perspectives. Therefore, in this paper, we advocate that stakeholders in AI risk management should be aware of the nuances, synergies, and interplay between safety and security, and unambiguously take into account the perspectives of both disciplines in order to devise mostly effective and holistic risk mitigation approaches. Unfortunately, this vision is often obfuscated, as the definitions of the basic concepts of "safety" and "security" themselves are often inconsistent and lack consensus across communities. With AI risk management being increasingly cross-disciplinary, this issue is particularly salient. In light of this conceptual challenge, we introduce a unified reference framework to clarify the differences and interplay between AI safety and AI security, aiming to facilitate a shared understanding and effective collaboration across communities.

cs.CR↗

arXiv2024

Embodied LLM Agents Learn to Cooperate in Organized Teams

Large Language Models (LLMs) have emerged as integral tools for reasoning, planning, and decision-making, drawing upon their extensive world knowledge and proficiency in language-related tasks. LLMs thus hold tremendous potential for natural language interaction within multi-agent systems to foster cooperation. However, LLM agents tend to over-report and comply with any instruction, which may result in information redundancy and confusion in multi-agent cooperation. Inspired by human organizations, this paper introduces a framework that imposes prompt-based organization structures on LLM agents to mitigate these problems. Through a series of experiments with embodied LLM agents and human-agent collaboration, our results highlight the impact of designated leadership on team efficiency, shedding light on the leadership qualities displayed by LLM agents and their spontaneous cooperative behaviors. Further, we harness the potential of LLMs to propose enhanced organizational prompts, via a Criticize-Reflect process, resulting in novel organization structures that reduce communication costs and enhance team efficiency.

cs.AI↗

arXiv2024

Diffusion Model for Data-Driven Black-Box Optimization

Generative AI has redefined artificial intelligence, enabling the creation of innovative content and customized solutions that drive business practices into a new era of efficiency and creativity. In this paper, we focus on diffusion models, a powerful generative AI technology, and investigate their potential for black-box optimization over complex structured variables. Consider the practical scenario where one wants to optimize some structured design in a high-dimensional space, based on massive unlabeled data (representing design variables) and a small labeled dataset. We study two practical types of labels: 1) noisy measurements of a real-valued reward function and 2) human preference based on pairwise comparisons. The goal is to generate new designs that are near-optimal and preserve the designed latent structures. Our proposed method reformulates the design optimization problem into a conditional sampling problem, which allows us to leverage the power of diffusion models for modeling complex distributions. In particular, we propose a reward-directed conditional diffusion model, to be trained on the mixed data, for sampling a near-optimal solution conditioned on high predicted rewards. Theoretically, we establish sub-optimality error bounds for the generated designs. The sub-optimality gap nearly matches the optimal guarantee in off-policy bandits, demonstrating the efficiency of reward-directed diffusion models for black-box optimization. Moreover, when the data admits a low-dimensional latent subspace structure, our model efficiently generates high-fidelity designs that closely respect the latent structure. We provide empirical experiments validating our model in decision-making and content-creation tasks.

cs.LG↗

arXiv2024

Density-dependent quark mean-field model for nuclear matter and neutron stars

We develop a density-dependent quark mean-field (DDQMF) model to study the properties of nuclear matter and neutron stars, where the coupling strength between $σ$ meson and nucleon is generated by the degree of freedom of quarks, while other meson coupling constants are regarded as density-dependent ones. Two values for the nucleon effective mass, $M^*_{N0}/M_N=0.556,~0.70$ at the saturation density are chosen based on the consideration of the core-collapse supernova simulation and finite nuclei when the meson-nucleon coupling constants are fixed. We find that the equation of state (EOS) of nuclear matter, the symmetry energy, the mass-radius relations, and the tidal deformabilities of neutron stars with larger nucleon effective mass are more sensitive to the skewness coefficient $J_0$. The EOSs with $M^*_{N0}/M_N=0.70$ are softer when the skewness coefficient $J_0=-800$ MeV. However, the maximum masses of the neutron star can be around $2.32M_\odot$ with $J_0=400$ MeV regardless of the value of the nucleon effective mass. By manipulating the coupling strength of the isovector meson to generate different slopes of symmetry energy, we construct the neutron star EOSs that can satisfy the different variables from the simultaneous mass-radius measurements of PSR J0030+0451, PSR J0740+6620 by the NICER collaboration, the mass-radius relations of HESS J1731-347, and the radius constraints from the gravitational-wave signal GW170817 in the framework of DDQMF model. At the same time, most of these constructed EOSs can also satisfy the constraints of the tidal deformability from GW170817 event.

nucl-th↗

arXiv2023

A 5' UTR Language Model for Decoding Untranslated Regions of mRNA and Function Predictions

The 5' UTR, a regulatory region at the beginning of an mRNA molecule, plays a crucial role in regulating the translation process and impacts the protein expression level. Language models have showcased their effectiveness in decoding the functions of protein and genome sequences. Here, we introduced a language model for 5' UTR, which we refer to as the UTR-LM. The UTR-LM is pre-trained on endogenous 5' UTRs from multiple species and is further augmented with supervised information including secondary structure and minimum free energy. We fine-tuned the UTR-LM in a variety of downstream tasks. The model outperformed the best-known benchmark by up to 42% for predicting the Mean Ribosome Loading, and by up to 60% for predicting the Translation Efficiency and the mRNA Expression Level. The model also applies to identifying unannotated Internal Ribosome Entry Sites within the untranslated region and improves the AUPR from 0.37 to 0.52 compared to the best baseline. Further, we designed a library of 211 novel 5' UTRs with high predicted values of translation efficiency and evaluated them via a wet-lab assay. Experiment results confirmed that our top designs achieved a 32.5% increase in protein production level relative to well-established 5' UTR optimized for therapeutics.

cs.LG↗

arXiv2023

A Variational Auto-Encoder Enabled Multi-Band Channel Prediction Scheme for Indoor Localization

Indoor localization is getting increasing demands for various cutting-edged technologies, like Virtual/Augmented reality and smart home. Traditional model-based localization suffers from significant computational overhead, so fingerprint localization is getting increasing attention, which needs lower computation cost after the fingerprint database is built. However, the accuracy of indoor localization is limited by the complicated indoor environment which brings the multipath signal refraction. In this paper, we provided a scheme to improve the accuracy of indoor fingerprint localization from the frequency domain by predicting the channel state information (CSI) values from another transmitting channel and spliced the multi-band information together to get more precise localization results. We tested our proposed scheme on COST 2100 simulation data and real time orthogonal frequency division multiplexing (OFDM) WiFi data collected from an office scenario.

eess.SP↗

arXiv2023

Simulating Chern insulators on a superconducting quantum processor

The quantum Hall effect, fundamental in modern condensed matter physics, continuously inspires new theories and predicts emergent phases of matter. Here we experimentally demonstrate three types of Chern insulators with synthetic dimensions on a programable 30-qubit-ladder superconducting processor. We directly measure the band structures of the 2D Chern insulator along synthetic dimensions with various configurations of Aubry-André-Harper chains and observe dynamical localisation of edge excitations. With these two signatures of topology, our experiments implement the bulk-edge correspondence in the synthetic 2D Chern insulator. Moreover, we simulate two different bilayer Chern insulators on the ladder-type superconducting processor. With the same and opposite periodically modulated on-site potentials for two coupled chains, we simulate topologically nontrivial edge states with zero Hall conductivity and a Chern insulator with higher Chern numbers, respectively. Our work shows the potential of using superconducting qubits for investigating different intriguing topological phases of quantum matter.

quant-ph↗

arXiv2023

Visual Adversarial Examples Jailbreak Aligned Large Language Models

Recently, there has been a surge of interest in integrating vision into Large Language Models (LLMs), exemplified by Visual Language Models (VLMs) such as Flamingo and GPT-4. This paper sheds light on the security and safety implications of this trend. First, we underscore that the continuous and high-dimensional nature of the visual input makes it a weak link against adversarial attacks, representing an expanded attack surface of vision-integrated LLMs. Second, we highlight that the versatility of LLMs also presents visual attackers with a wider array of achievable adversarial objectives, extending the implications of security failures beyond mere misclassification. As an illustration, we present a case study in which we exploit visual adversarial examples to circumvent the safety guardrail of aligned LLMs with integrated vision. Intriguingly, we discover that a single visual adversarial example can universally jailbreak an aligned LLM, compelling it to heed a wide range of harmful instructions that it otherwise would not) and generate harmful content that transcends the narrow scope of a `few-shot' derogatory corpus initially employed to optimize the adversarial example. Our study underscores the escalating adversarial risks associated with the pursuit of multimodality. Our findings also connect the long-studied adversarial vulnerabilities of neural networks to the nascent field of AI alignment. The presented attack suggests a fundamental adversarial challenge for AI alignment, especially in light of the emerging trend toward multimodality in frontier foundation models.

cs.CR↗