arXiv subjects

Chao Zhang

Publications and source records attributed to Chao Zhang.

At least 451 records · Page 25Linked to original sources

arXiv2024

Ionic Selectivity of Nanopores: Comparison among Cases under the Hydrostatic Pressure, Electric Field, and Concentration Gradient

The ionic selectivity of nanopores is crucial for the energy conversion based on nanoporous membranes. It can be significantly affected by various parameters of nanopores and the applied fields driving ions through porous membranes. Here, with finite element simulations, the selective transport of ions through nanopores is systematically investigated under three common fields, i.e. the electric field (V), hydrostatic pressure (p), and concentration gradient (C). For negatively charged nanopores, through the quantitative comparison of the cation selectivity (t+) under the three fields, the cation selectivity of nanopores follows the order of t+V > t+c > t+p. This is due to the transport characteristics of cations and anions through the nanopores. Because of the strong transport of counterions in electric double layers under electric fields and concentration gradients, the nanopore exhibits a relatively higher selectivity to counterions. We also explored the modulation of t+ on the properties of nanopores and solutions. Under all three fields, t+ is directly proportional to the pore length and surface charge density, and inversely correlated to the pore diameter and salt concentration. Under both the electric field and hydrostatic pressure, t+ has almost no dependence on the applied field strength or ion species, which can affect t+ in the case of the concentration gradient. Our results provide detailed insights into the comparison and regulation of ionic selectivity of nanopores under three fields which can be useful for the design of high-performance devices for energy conversion based on nanoporous membranes.

cond-mat.soft↗

arXiv2024

Modulation of ionic current rectification in short bipolar nanopores

Bipolar nanopores, with asymmetric charge distributions, can induce significant ionic current rectification (ICR) at ultra-short lengths, finding potential applications in nanofluidic devices, energy conversion, and other related fields. Here, with simulations, we investigated the characteristics of ion transport and modulation of ICR inside bipolar nanopores. With bipolar nanopores of half-positive and half-negative surfaces, the most significant ICR phenomenon appears at various concentrations. In these cases, ICR ratios are independent of electrolyte types. In other cases where nanopores have oppositely charged surfaces in different lengths, ICR ratios are related to the mobility of anions and cations. The pore length and surface charge density can enhance ICR. As the pore length increases, ICR ratios first increase and then approach their saturation which is determined by the surface charge density. External surface charges of nanopores can promote the ICR phenomenon mainly due to the enhancement of ion enrichment inside nanopores by external surface conductance. The effective width of exterior charged surfaces under various conditions is also explored, which is inversely proportional to the pore length and salt concentration, and linearly related to the pore diameter, surface charge density, and applied voltage. Our results may provide guidance for the design of bipolar porous membranes.

physics.chem-ph↗

arXiv2024

Two-Step Offline Preference-Based Reinforcement Learning with Constrained Actions

Preference-based reinforcement learning (PBRL) in the offline setting has succeeded greatly in industrial applications such as chatbots. A two-step learning framework where one applies a reinforcement learning step after a reward modeling step has been widely adopted for the problem. However, such a method faces challenges from the risk of reward hacking and the complexity of reinforcement learning. To overcome the challenge, our insight is that both challenges come from the state-actions not supported in the dataset. Such state-actions are unreliable and increase the complexity of the reinforcement learning problem at the second step. Based on the insight, we develop a novel two-step learning method called PRC: preference-based reinforcement learning with constrained actions. The high-level idea is to limit the reinforcement learning agent to optimize over a constrained action space that excludes the out-of-distribution state-actions. We empirically verify that our method has high learning efficiency on various datasets in robotic control environments.

cs.LG↗

arXiv2024

HYDRA: Model Factorization Framework for Black-Box LLM Personalization

Personalization has emerged as a critical research area in modern intelligent systems, focusing on mining users' behavioral history and adapting to their preferences for delivering tailored experiences. Despite the remarkable few-shot capabilities exhibited by black-box large language models (LLMs), the inherent opacity of their model parameters presents significant challenges in aligning the generated output with individual expectations. Existing solutions have primarily focused on prompt design to incorporate user-specific profiles and behaviors; however, such approaches often struggle to generalize effectively due to their inability to capture shared knowledge among all users. To address these challenges, we propose HYDRA, a model factorization framework that captures both user-specific behavior patterns from historical data and shared general knowledge among all users to deliver personalized generation. In order to capture user-specific behavior patterns, we first train a reranker to prioritize the most useful information from top-retrieved relevant historical records. By combining the prioritized history with the corresponding query, we train an adapter to align the output with individual user-specific preferences, eliminating the reliance on access to inherent model parameters of black-box LLMs. Both the reranker and the adapter can be decomposed into a base model with multiple user-specific heads, resembling a hydra. The base model maintains shared knowledge across users, while the multiple personal heads capture user-specific preferences. Experimental results demonstrate that HYDRA outperforms existing state-of-the-art prompt-based methods by an average relative improvement of 9.01% across five diverse personalization tasks in the LaMP benchmark. Our implementation is available at https://github.com/night-chen/HYDRA.

cs.CL↗

arXiv2024

Quasi-Normal Modes of Loop Quantum Black Holes Formed from Gravitational Collapse

In this paper, we study the quasi-normal modes (QNMs) of a scalar field in the background of a large class of quantum black holes that can be formed from gravitational collapse of a dust fluid in the framework of effective loop quantum gravity. The loop quantum black holes (LQBHs) are characterized by three free parameters, one of which is the mass parameter, while the other two are purely due to quantum geometric effects. Among these two quantum parameters, one is completely fixed by black hole thermodynamics and its effects are negligible for macroscopic black holes, while the second parameter is completely free (in principle). In the studies of the QNMs of such LQBHs, we pay particular attention to the difference of the QNMs between LQBHs and classical ones, so that they can be observed for the current and forthcoming gravitational wave observations, whereby place the LQBH theory directly under the test of observations.

gr-qc↗

arXiv2024

A new approach to N-doped di-molybdenum carbide with enhanced superconductivity via Urea

Chemical doping is a critical factor in the development of new superconductors or optimizing the superconducting transition temperature (Tc) of the parent superconducting materials. Herein, a new simple urea approach is developed to synthesize the N-doped alfa-Mo2C. Benefiting from the simple urea method, a broad superconducting dome is found in the Mo2C1-xNx compositions. XRD results show that the structure of alfa-Mo2C remains unchanged and that there is a variation of lattice parameters with nitrogen doping. Resistivity, magnetic susceptibility, and heat capacity measurement results confirm that the superconducting transition temperature (Tc) was strongly increased from 2.68 K (x = 0) to 7.05 K (x = 0.49). First-principles calculations and our analysis indicate that increasing nitrogen doping leads to a rise in the density of states at the Fermi level and doping-induced phonon softening, which enhances electron-phonon coupling. This results in an increase in Tc and a sharp rise in the upper critical field. Our findings provide a promising strategy for fabricating transition metal carbonitrides and provide a material platform for further study of the superconductivity of transition metal carbides.

cond-mat.supr-con↗

arXiv2024

Microscopic Ampère current-current interaction

With the rapid development of modern measurement techniques, the energy resolution of $1 \, meV$ can now be easily obtained. Generally, the driving mechanisms of the physical, chemical or biological processes of the matters or the living organisms on Earth at about $1 \, meV$ energy scale are assumed to stem from the fundamental microscopic Coulomb interaction, its various reduced ones and the relativistic corrections. In this article, by using a path integral approach on a non-relativistic quantum electrodynamics theory, we show that there is another fundamental microscopic electromagnetic interaction at this energy scale, the microscopic Ampère current-current interaction. It has time-dependent dynamical feature and can be the driving interaction of the physical, chemical or biological processes at about $1\, meV$ energy scale. A new Ampère-type exchange spin interaction is also found with a magnitude about $10^{-4}$ of the well-known Heisenberg exchange spin interaction.

cond-mat.str-el↗

arXiv2024

Large Language Model Based Generative Error Correction: A Challenge and Baselines for Speech Recognition, Speaker Tagging, and Emotion Recognition

Given recent advances in generative AI technology, a key question is how large language models (LLMs) can enhance acoustic modeling tasks using text decoding results from a frozen, pretrained automatic speech recognition (ASR) model. To explore new capabilities in language modeling for speech processing, we introduce the generative speech transcription error correction (GenSEC) challenge. This challenge comprises three post-ASR language modeling tasks: (i) post-ASR transcription correction, (ii) speaker tagging, and (iii) emotion recognition. These tasks aim to emulate future LLM-based agents handling voice-based interfaces while remaining accessible to a broad audience by utilizing open pretrained language models or agent-based APIs. We also discuss insights from baseline evaluations, as well as lessons learned for designing future evaluations.

cs.CL↗

arXiv2024

Enhancing Multimodal LLM for Detailed and Accurate Video Captioning using Multi-Round Preference Optimization

Videos contain a wealth of information, and generating detailed and accurate descriptions in natural language is a key aspect of video understanding. In this paper, we present video-SALMONN 2, an advanced audio-visual large language model (LLM) with low-rank adaptation (LoRA) designed for enhanced video (with paired audio) captioning through directed preference optimization (DPO). We propose new metrics to evaluate the completeness and accuracy of video descriptions, which are optimized using DPO. To further improve training, we introduce a novel multi-round DPO (mrDPO) approach, which involves periodically updating the DPO reference model, merging and re-initializing the LoRA module as a proxy for parameter updates after each training round (1,000 steps), and incorporating guidance from ground-truth video captions to stabilize the process. To address potential catastrophic forgetting of non-captioning abilities due to mrDPO, we propose rebirth tuning, which finetunes the pre-DPO LLM by using the captions generated by the mrDPO-trained model as supervised labels. Experiments show that mrDPO significantly enhances video-SALMONN 2's captioning accuracy, reducing global and local error rates by 40\% and 20\%, respectively, while decreasing the repetition rate by 35\%. The final video-SALMONN 2 model, with just 7 billion parameters, surpasses leading models such as GPT-4o and Gemini-1.5-Pro in video captioning tasks, while maintaining competitive performance to the state-of-the-art on widely used video question-answering benchmark among models of similar size. Upon acceptance, we will release the code, model checkpoints, and training and test data. Demos are available at \href{https://video-salmonn-2.github.io}{https://video-salmonn-2.github.io}.

cs.CV↗

arXiv2024

Characterizing Quantum Codes via the Coefficients in Knill-Laflamme Conditions

Quantum error correction (QEC) is essential for protecting quantum information against noise, yet understanding the structure of the Knill-Laflamme (KL) coefficients $λ_{ij}$ from the condition $PE_i^\dagger E_j P = λ_{ij} P$ remains challenging, particularly for nonadditive codes. In this work, we introduce the signature vector $\vecλ(P)$, composed of the off-diagonal KL coefficients $λ_{ij}$, where each coefficient corresponds to equivalence classes of errors counted only once. We define its Euclidean norm $λ^*(P)$ as a scalar measure representing the total strength of error correlations within the code subspace defined by the projector $P$. We parameterize $P$ on a Stiefel manifold and formulate an optimization problem based on the KL conditions to systematically explore possible values of $λ^*$. Moreover, we show that, for $((n,K,d))$ codes, $λ^*$ is invariant under local unitary transformations. Applying our approach to the $((6, 2, 3))$ quantum code, we find that $λ^*_{\text{min}} = \sqrt{0.6}$ and $λ^*_{\text{max}} = 1$, with $λ^* = 1$ corresponding to a known degenerate stabilizer code. We construct continuous families of new nonadditive codes parameterized by vectors in $\mathbb{R}^5$, with $λ^*$ varying over the interval $[\sqrt{0.6}, 1]$. For the $((7, 2, 3))$ code, we identify $λ^*_{\text{min}} = 0$ (corresponding to the non-degenerate Steane code) and $λ^*_{\text{max}} = \sqrt{7}$ (corresponding to the permutation-invariant code by Pollatsek and Ruskai), and we demonstrate continuous paths connecting these extremes via cyclic codes characterized solely by $λ^*$. Our findings provide new insights into the structure of quantum codes, advance the theoretical foundations of QEC, and open new avenues for investigating intricate relationships between code subspaces and error correlations.

quant-ph↗

arXiv2024

PACER: A Fully Push-forward-based Distributional Reinforcement Learning Algorithm

In this paper, we propose the first fully push-forward-based distributional reinforcement learning algorithm, named PACER, which consists of a distributional critic, a stochastic actor and a sample-based encourager. Specifically, the push-forward operator is leveraged in both the critic and actor to model the return distributions and stochastic policies respectively, enabling them with equal modeling capability and thus enhancing the synergetic performance. Since it is infeasible to obtain the density function of the push-forward policies, novel sample-based regularizers are integrated in the encourager to incentivize efficient exploration and alleviate the risk of trapping into local optima. Moreover, a sample-based stochastic utility value policy gradient is established for the push-forward policy update, which circumvents the explicit demand of the policy density function in existing REINFORCE-based stochastic policy gradient. As a result, PACER fully utilizes the modeling capability of the push-forward operator and is able to explore a broader class of the policy space, compared with limited policy classes used in existing distributional actor critic algorithms (i.e. Gaussians). We validate the critical role of each component in our algorithm with extensive empirical studies. Experimental results demonstrate the superiority of our algorithm over the state-of-the-art.

cs.LG↗

arXiv2024

BELM: Bidirectional Explicit Linear Multi-step Sampler for Exact Inversion in Diffusion Models

The inversion of diffusion model sampling, which aims to find the corresponding initial noise of a sample, plays a critical role in various tasks. Recently, several heuristic exact inversion samplers have been proposed to address the inexact inversion issue in a training-free manner. However, the theoretical properties of these heuristic samplers remain unknown and they often exhibit mediocre sampling quality. In this paper, we introduce a generic formulation, \emph{Bidirectional Explicit Linear Multi-step} (BELM) samplers, of the exact inversion samplers, which includes all previously proposed heuristic exact inversion samplers as special cases. The BELM formulation is derived from the variable-stepsize-variable-formula linear multi-step method via integrating a bidirectional explicit constraint. We highlight this bidirectional explicit constraint is the key of mathematically exact inversion. We systematically investigate the Local Truncation Error (LTE) within the BELM framework and show that the existing heuristic designs of exact inversion samplers yield sub-optimal LTE. Consequently, we propose the Optimal BELM (O-BELM) sampler through the LTE minimization approach. We conduct additional analysis to substantiate the theoretical stability and global convergence property of the proposed optimal sampler. Comprehensive experiments demonstrate our O-BELM sampler establishes the exact inversion property while achieving high-quality sampling. Additional experiments in image editing and image interpolation highlight the extensive potential of applying O-BELM in varying applications.

cs.CV↗

arXiv2024

Retrieve-Plan-Generation: An Iterative Planning and Answering Framework for Knowledge-Intensive LLM Generation

Despite the significant progress of large language models (LLMs) in various tasks, they often produce factual errors due to their limited internal knowledge. Retrieval-Augmented Generation (RAG), which enhances LLMs with external knowledge sources, offers a promising solution. However, these methods can be misled by irrelevant paragraphs in retrieved documents. Due to the inherent uncertainty in LLM generation, inputting the entire document may introduce off-topic information, causing the model to deviate from the central topic and affecting the relevance of the generated content. To address these issues, we propose the Retrieve-Plan-Generation (RPG) framework. RPG generates plan tokens to guide subsequent generation in the plan stage. In the answer stage, the model selects relevant fine-grained paragraphs based on the plan and uses them for further answer generation. This plan-answer process is repeated iteratively until completion, enhancing generation relevance by focusing on specific topics. To implement this framework efficiently, we utilize a simple but effective multi-task prompt-tuning method, enabling the existing LLMs to handle both planning and answering. We comprehensively compare RPG with baselines across 5 knowledge-intensive generation tasks, demonstrating the effectiveness of our approach.

cs.CL↗

arXiv2024

Nonlocal photonic quantum gates over 7.0 km

Quantum networks provide a prospective paradigm to connect separated quantum nodes, which relies on the distribution of long-distance entanglement and active feedforward control of qubits between remote nodes. Such approaches can be utilized to construct nonlocal quantum gates, forming building blocks for distributed quantum computing and other novel quantum applications. However, these gates have only been realized within single nodes or between nodes separated by a few tens of meters, limiting the ability to harness computing resources in large-scale quantum networks. Here, we demonstrate nonlocal photonic quantum gates between two nodes spatially separated by 7.0 km using stationary qubits based on multiplexed quantum memories, flying qubits at telecom wavelengths, and active feedforward control based on field-deployed fibers. Furthermore, we illustrate quantum parallelism by implementing the Deutsch-Jozsa algorithm and the quantum phase estimation algorithm between the two remote nodes. These results represent a proof-of-principle demonstration of quantum gates over metropolitan-scale distances and lay the foundation for the construction of large-scale distributed quantum networks relying on existing fiber channels.

quant-ph↗

arXiv2024

BMRetriever: Tuning Large Language Models as Better Biomedical Text Retrievers

Developing effective biomedical retrieval models is important for excelling at knowledge-intensive biomedical tasks but still challenging due to the deficiency of sufficient publicly annotated biomedical data and computational resources. We present BMRetriever, a series of dense retrievers for enhancing biomedical retrieval via unsupervised pre-training on large biomedical corpora, followed by instruction fine-tuning on a combination of labeled datasets and synthetic pairs. Experiments on 5 biomedical tasks across 11 datasets verify BMRetriever's efficacy on various biomedical applications. BMRetriever also exhibits strong parameter efficiency, with the 410M variant outperforming baselines up to 11.7 times larger, and the 2B variant matching the performance of models with over 5B parameters. The training data and model checkpoints are released at \url{https://huggingface.co/BMRetriever} to ensure transparency, reproducibility, and application to new domains.

cs.CL↗

arXiv2024

LoRC: Low-Rank Compression for LLMs KV Cache with a Progressive Compression Strategy

The Key-Value (KV) cache is a crucial component in serving transformer-based autoregressive large language models (LLMs), enabling faster inference by storing previously computed KV vectors. However, its memory consumption scales linearly with sequence length and batch size, posing a significant bottleneck in LLM deployment. Existing approaches to mitigate this issue include: (1) efficient attention variants integrated in upcycling stages, which requires extensive parameter tuning thus unsuitable for pre-trained LLMs; (2) KV cache compression at test time, primarily through token eviction policies, which often overlook inter-layer dependencies and can be task-specific. This paper introduces an orthogonal approach to KV cache compression. We propose a low-rank approximation of KV weight matrices, allowing for plug-in integration with existing transformer-based LLMs without model retraining. To effectively compress KV cache at the weight level, we adjust for layerwise sensitivity and introduce a progressive compression strategy, which is supported by our theoretical analysis on how compression errors accumulate in deep networks. Our method is designed to function without model tuning in upcycling stages or task-specific profiling in test stages. Extensive experiments with LLaMA models ranging from 8B to 70B parameters across various tasks show that our approach significantly reduces the GPU memory footprint while maintaining performance.

cs.LG↗

arXiv2024

Coincidence detection probability of $(γ, 2e)$ photoemission measurement

In the study of strongly correlated electrons, one of the challenging core tasks is to develop the potential techniques for direct detection of the many-body correlations of strongly correlated electrons. The $(γ, 2e)$ photoemission technique has been developed to investigate the two-body correlations of the target correlated electrons. In this article, we will focus on this technique for the correlated electrons near the Fermi energy in condensed matter. The coincidence detection probability of the two emitted electrons in the $(γ, 2e)$ photoemission measurement is shown to be relevant to a two-body Bethe-Salpeter wave function, which describes the dynamical two-body correlations of the target correlated electrons near the Fermi energy. As the coincidence detection probability involves an electron-electron interaction matrix element, the arbitrary momentum and/or energy transfer due to this electron-electron interaction makes the $(γ, 2e)$ photoemission technique fail to reveal the inner-pair structures of the two-body Bethe-Salpeter wave function. However, the center-of-mass momentum and energy of the two-body Bethe-Salpeter wave function can be distinctly resolved. Thus, the $(γ, 2e)$ photoemission technique can provide the center-of-mass physics of the two-body correlations of the target correlated electrons. It will be one potential technique to study the center-of-mass physics of the Cooper pairs in superconductor.

cond-mat.supr-con↗

arXiv2024

Enhancing heat transfer in X-ray tube by van der heterostructures-based thermionic emission

Van der Waals (vdW) heterostructures have attracted much attention due to their distinctive optical, electrical, and thermal properties, demonstrating promising potential in areas such as photocatalysis, ultrafast photonics, and free electron radiation devices. Particularly, they are promising platforms for studying thermionic emission. Here, we illustrate that using vdW heterostructure-based thermionic emission can enhance heat transfer in vacuum devices. As a proof of concept, we demonstrate that this approach offers a promising solution to the long-standing overheating issue in X-ray tubes. Specifically, we show that the saturated target temperature of a 2000 W X-ray tube can be reduced from around 1200 celsius to 490 celsius. Additionally, our study demonstrates that by reducing the height of the Schottky barrier formed in the vdW heterostructures, the thermionic cooling performance can be enhanced. Our findings pave the way for the development of high-power X-ray tubes.

physics.app-ph↗