Searcharxiv⌕ Search

arXiv subjects

Zijian Wang

Publications and source records attributed to Zijian Wang.

At least 127 records · Page 7Linked to original sources
arXiv2025

Intrinsic Mixed-state Topological Order

Decoherence is a major obstacle to the preparation of topological order in noisy intermediate-scale quantum devices. Here, we show that decoherence can also give rise to new types of topological order. Specifically, we construct concrete examples by proliferating fermionic anyons in the toric code via local quantum channels. The resulting mixed states retain long-range entanglement, which manifests in the nonzero topological entanglement negativity, though the topological quantum memory is destroyed by decoherence. By comparison with the gapless spin liquid in pure states, we show that the identified states represent a novel intrinsic mixed-state topological order, which has no counterpart in pure states. Through the lens of quantum anomalies of 1-form symmetries, we then provide general constructions of intrinsic mixed-state topological order and reveal the existence of nonbosonic deconfined anyons as another key feature of these novel phases. The extended meaning and characterization of deconfined excitations and their statistics in mixed states are clarified. Moreover, when these deconfined anyons have nontrivial braiding statistics, we prove that the mixed states cannot be prepared via finite-depth local quantum channels from any bipartite separable states. We further demonstrate our construction using the decohered Kitaev honeycomb model and the decohered double-semion model. In the latter case, a surprising scenario arises where decoherence gives rise to additional types of deconfined anyons.

quant-ph↗
arXiv2025

Learn A Flexible Exploration Model for Parameterized Action Markov Decision Processes

Hybrid action models are widely considered an effective approach to reinforcement learning (RL) modeling. The current mainstream method is to train agents under Parameterized Action Markov Decision Processes (PAMDPs), which performs well in specific environments. Unfortunately, these models either exhibit drastic low learning efficiency in complex PAMDPs or lose crucial information in the conversion between raw space and latent space. To enhance the learning efficiency and asymptotic performance of the agent, we propose a model-based RL (MBRL) algorithm, FLEXplore. FLEXplore learns a parameterized-action-conditioned dynamics model and employs a modified Model Predictive Path Integral control. Unlike conventional MBRL algorithms, we carefully design the dynamics loss function and reward smoothing process to learn a loose yet flexible model. Additionally, we use the variational lower bound to maximize the mutual information between the state and the hybrid action, enhancing the exploration effectiveness of the agent. We theoretically demonstrate that FLEXplore can reduce the regret of the rollout trajectory through the Wasserstein Metric under given Lipschitz conditions. Our empirical results on several standard benchmarks show that FLEXplore has outstanding learning efficiency and asymptotic performance compared to other baselines.

cs.LG↗
arXiv2024

Walk Wisely on Graph: Knowledge Graph Reasoning with Dual Agents via Efficient Guidance-Exploration

Recent years, multi-hop reasoning has been widely studied for knowledge graph (KG) reasoning due to its efficacy and interpretability. However, previous multi-hop reasoning approaches are subject to two primary shortcomings. First, agents struggle to learn effective and robust policies at the early phase due to sparse rewards. Second, these approaches often falter on specific datasets like sparse knowledge graphs, where agents are required to traverse lengthy reasoning paths. To address these problems, we propose a multi-hop reasoning model with dual agents based on hierarchical reinforcement learning (HRL), which is named FULORA. FULORA tackles the above reasoning challenges by eFficient GUidance-ExpLORAtion between dual agents. The high-level agent walks on the simplified knowledge graph to provide stage-wise hints for the low-level agent walking on the original knowledge graph. In this framework, the low-level agent optimizes a value function that balances two objectives: (1) maximizing return, and (2) integrating efficient guidance from the high-level agent. Experiments conducted on three real-word knowledge graph datasets demonstrate that FULORA outperforms RL-based baselines, especially in the case of long-distance reasoning.

cs.AI↗
arXiv2024

Color-Oriented Redundancy Reduction in Dataset Distillation

Dataset Distillation (DD) is designed to generate condensed representations of extensive image datasets, enhancing training efficiency. Despite recent advances, there remains considerable potential for improvement, particularly in addressing the notable redundancy within the color space of distilled images. In this paper, we propose AutoPalette, a framework that minimizes color redundancy at the individual image and overall dataset levels, respectively. At the image level, we employ a palette network, a specialized neural network, to dynamically allocate colors from a reduced color space to each pixel. The palette network identifies essential areas in synthetic images for model training and consequently assigns more unique colors to them. At the dataset level, we develop a color-guided initialization strategy to minimize redundancy among images. Representative images with the least replicated color patterns are selected based on the information gain. A comprehensive performance study involving various datasets and evaluation scenarios is conducted, demonstrating the superior performance of our proposed color-aware DD compared to existing DD methods. The code is available at \url{https://github.com/KeViNYuAn0314/AutoPalette}.

cs.CV↗
arXiv2024

LibEvolutionEval: A Benchmark and Study for Version-Specific Code Generation

Recent advancements in code completion models have primarily focused on local file contexts. However, these studies do not fully capture the complexity of real-world software development, which often requires the use of rapidly-evolving public libraries. To fill the gap, we introduce LibEvolutionEval, a detailed study requiring an understanding of library evolution to perform in-line code completion accurately. LibEvolutionEval provides a version-specific code-completion task comprised of eight libraries (torch, torchvision, scipy, pil, tqdm, pyyaml, matplotlib, and pandas) as they evolve over the year along with a detailed analysis of the evolution of two popular and well-maintained public libraries: PyTorch and Matplotlib. We evaluate popular public models and find that public library evolution significantly influences model performance. We explored mitigation methods by studying how retrieved version-specific library documentation and prompting can improve the model's capability in handling these fast-evolving packages, paving a promising future path in better handling fast-evolving libraries.

cs.SE↗
arXiv2024

CodeFort: Robust Training for Code Generation Models

Code generation models are not robust to small perturbations, which often lead to incorrect generations and significantly degrade the performance of these models. Although improving the robustness of code generation models is crucial to enhancing user experience in real-world applications, existing research efforts do not address this issue. To fill this gap, we propose CodeFort, a framework to improve the robustness of code generation models, generalizing a large variety of code perturbations to enrich the training data and enabling various robust training strategies, mixing data augmentation, batch augmentation, adversarial logits pairing, and contrastive learning, all carefully designed to support high-throughput training. Extensive evaluations show that we increase the average robust pass rates of baseline CodeGen models from 14.79 to 21.74. We notably decrease the robustness drop rate from 95.02% to 54.95% against code-syntax perturbations.

cs.SE↗
arXiv2024

Study of charged Lepton Flavor Violation in electron muon interactions

With the improvement of muon acceleration technology, it has received great interest to exploit high-energy muon beams for collision or target experiments. We investigate possible charged Lepton Flavor Violation (cLFV) processes mediated by an extra massive neutral gauge boson Zprime in electron muon interactions, either at a proposed electron muon collider or in a fixed target experiment with high-energy muon beam hitting electrons in the target. Based on Monte Carlo calculations and fast detector simulations, we study in detail our signal and possible backgrounds, giving the sensitivity results of cLFV signals at the 90% confidence level. Compared with current and prospective limits set by other experiments, electron muon interactions demonstrate significant advantages in the cLFV coupling strength sensitivity with tau in the final states. In addition, a special cLFV coupling combination, lambda_emu * lambda_mumu, can also be probed in our proposal.

hep-ex↗
arXiv2024

Probing charged lepton flavor violation in an economical muon on-target experiment

This work proposes a new yet economical experiment to probe the charged lepton flavor violation (CLFV) process mediated by an extra massive neutron gauge boson $Z^\prime$ beyond the standard model, by extending a recently proposed muon dark matter project in the Peking University Muon (PKMuon) Experiment. The devices used originally for light mass dark matter direct detection are easily adaptable to search for the $μ^+e^- \to μ^+μ^-$ CLFV process leveraging the large-area, high-precision muon tracking and tomography system sandwiching a fixed target the incoming muons scatter off. The $μ^+μ^-$ final state signal studied in this work can be uniquely sensitive to specific CLFV parameter combinations, such as the couplings between $Z^\prime$, electron and muon, or $Z^\prime$ and two muons. Prospected results are obtained through detailed detector simulation for the proposal interfacing with a muon beam with energy at tens of $\mathrm{GeV}$ and a flux of $10^6\ \mathrm{s^{-1}}$. Based mainly on angular information of the incoming and outgoing particles, the expected upper limit at 95\% confidence level on the coupling coefficients $λ_{eμ}λ_{μμ}$ is able to reach $10^{-5}$ with, for example, $Z^\prime$ mass $0.25\ \mathrm{GeV}$, for a one year's run.

hep-ex↗
arXiv2024

Learning Code Preference via Synthetic Evolution

Large Language Models (LLMs) have recently demonstrated remarkable coding capabilities. However, assessing code generation based on well-formed properties and aligning it with developer preferences remains challenging. In this paper, we explore two key questions under the new challenge of code preference learning: (i) How do we train models to predict meaningful preferences for code? and (ii) How do human and LLM preferences align with verifiable code properties and developer code tastes? To this end, we propose CodeFavor, a framework for training pairwise code preference models from synthetic evolution data, including code commits and code critiques. To evaluate code preferences, we introduce CodePrefBench, a benchmark comprising 1364 rigorously curated code preference tasks to cover three verifiable properties-correctness, efficiency, and security-along with human preference. Our evaluation shows that CodeFavor holistically improves the accuracy of model-based code preferences by up to 28.8%. Meanwhile, CodeFavor models can match the performance of models with 6-9x more parameters while being 34x more cost-effective. We also rigorously validate the design choices in CodeFavor via a comprehensive set of controlled experiments. Furthermore, we discover the prohibitive costs and limitations of human-based code preference: despite spending 23.4 person-minutes on each task, 15.1-40.3% of tasks remain unsolved. Compared to model-based preference, human preference tends to be more accurate under the objective of code correctness, while being sub-optimal for non-functional objectives.

cs.LG↗
arXiv2024

Hawk: An Efficient NALM System for Accurate Low-Power Appliance Recognition

Non-intrusive Appliance Load Monitoring (NALM) aims to recognize individual appliance usage from the main meter without indoor sensors. However, existing systems struggle to balance dataset construction efficiency and event/state recognition accuracy, especially for low-power appliance recognition. This paper introduces Hawk, an efficient and accurate NALM system that operates in two stages: dataset construction and event recognition. In the data construction stage, we efficiently collect a balanced and diverse dataset, HawkDATA, based on balanced Gray code and enable automatic data annotations via a sampling synchronization strategy called shared perceptible time. During the event recognition stage, our algorithm integrates steady-state differential pre-processing and voting-based post-processing for accurate event recognition from the aggregate current. Experimental results show that HawkDATA takes only 1/71.5 of the collection time to collect 6.34x more appliance state combinations than the baseline. In HawkDATA and a widely used dataset, Hawk achieves an average F1 score of 93.94% for state recognition and 97.07% for event recognition, which is a 47. 98% and 11. 57% increase over SOTA algorithms. Furthermore, selected appliance subsets and the model trained from HawkDATA are deployed in two real-world scenarios with many unknown background appliances. The average F1 scores of event recognition are 96.02% and 94.76%. Hawk's source code and HawkDATA are accessible at https://github.com/WZiJ/SenSys24-Hawk.

eess.SP↗
arXiv2024

Open-CRB: Towards Open World Active Learning for 3D Object Detection

LiDAR-based 3D object detection has recently seen significant advancements through active learning (AL), attaining satisfactory performance by training on a small fraction of strategically selected point clouds. However, in real-world deployments where streaming point clouds may include unknown or novel objects, the ability of current AL methods to capture such objects remains unexplored. This paper investigates a more practical and challenging research task: Open World Active Learning for 3D Object Detection (OWAL-3D), aimed at acquiring informative point clouds with new concepts. To tackle this challenge, we propose a simple yet effective strategy called Open Label Conciseness (OLC), which mines novel 3D objects with minimal annotation costs. Our empirical results show that OLC successfully adapts the 3D detection model to the open world scenario with just a single round of selection. Any generic AL policy can then be integrated with the proposed OLC to efficiently address the OWAL-3D problem. Based on this, we introduce the Open-CRB framework, which seamlessly integrates OLC with our preliminary AL method, CRB, designed specifically for 3D object detection. We develop a comprehensive codebase for easy reproducing and future research, supporting 15 baseline methods (\textit{i.e.}, active learning, out-of-distribution detection and open world detection), 2 types of modern 3D detectors (\textit{i.e.}, one-stage SECOND and two-stage PV-RCNN) and 3 benchmark 3D datasets (\textit{i.e.}, KITTI, nuScenes and Waymo). Extensive experiments evidence that the proposed Open-CRB demonstrates superiority and flexibility in recognizing both novel and known classes with very limited labeling costs, compared to state-of-the-art baselines. Source code is available at \url{https://github.com/Luoyadan/CRB-active-3Ddet/tree/Open-CRB}.

cs.CV↗
arXiv2024

Overview of AI-Debater 2023: The Challenges of Argument Generation Tasks

In this paper we present the results of the AI-Debater 2023 Challenge held by the Chinese Conference on Affect Computing (CCAC 2023), and introduce the related datasets. We organize two tracks to handle the argumentative generation tasks in different scenarios, namely, Counter-Argument Generation (Track 1) and Claim-based Argument Generation (Track 2). Each track is equipped with its distinct dataset and baseline model respectively. In total, 32 competing teams register for the challenge, from which we received 11 successful submissions. In this paper, we will present the results of the challenge and a summary of the systems, highlighting commonalities and innovations among participating systems. Datasets and baseline models of the AI-Debater 2023 Challenge have been already released and can be accessed through the official website of the challenge.

cs.CL↗
arXiv2024

Locating and Extracting Relational Concepts in Large Language Models

Relational concepts are indeed foundational to the structure of knowledge representation, as they facilitate the association between various entity concepts, allowing us to express and comprehend complex world knowledge. By expressing relational concepts in natural language prompts, people can effortlessly interact with large language models (LLMs) and recall desired factual knowledge. However, the process of knowledge recall lacks interpretability, and representations of relational concepts within LLMs remain unknown to us. In this paper, we identify hidden states that can express entity and relational concepts through causal mediation analysis in fact recall processes. Our finding reveals that at the last token position of the input prompt, there are hidden states that solely express the causal effects of relational concepts. Based on this finding, we assume that these hidden states can be treated as relational representations and we can successfully extract them from LLMs. The experimental results demonstrate high credibility of the relational representations: they can be flexibly transplanted into other fact recall processes, and can also be used as robust entity connectors. Moreover, we also show that the relational representations exhibit significant potential for controllable fact recall through relation rewriting.

cs.CL↗
arXiv2024

Anomaly in open quantum systems and its implications on mixed-state quantum phases

In this paper, we develop a systematic approach to characterize the 't Hooft anomaly in open quantum systems. Owing to nontrivial couplings to the environment, symmetries in such systems manifest as either strong or weak type. By representing their symmetry transformation through superoperators, we incorporate them in a unified framework that enables a direct calculation of their anomalies. In the case where the full symmetry group is $K\times G$, with $K$ the strong symmetry and $G$ the weak symmetry, we find that anomalies of bosonic systems are classified by $H^{d+2}(K\times G,U(1))/H^{d+2}(G,U(1))$ in $d$ spatial dimensions. To illustrate the power of anomalies in open quantum systems, we generally prove that anomaly must lead to nontrivial mixed-state quantum phases as long as the weak symmetry is imposed. Analogous to the ``anomaly matching" condition ensuring nontrivial low-energy physics in closed systems, anomaly also guarantees nontrivial steady states and long-time dynamics for open quantum systems governed by Lindbladians. Notably, we identify a novel $(1+1)$-D mixed-state quantum phase that has no counterpart in closed systems, where the steady state shows no nontrivial correlation function in the bulk, but displays spontaneous symmetry breaking order on the boundary, which is enforced by anomalies. We further establish the general relations between mixed-state anomalies and such unconventional boundary correlation. Finally, we explore the generalization of the ``anomaly inflow" mechanism in open quantum systems. We construct $(1+1)$-D and $(2+1)$-D Lindbladians whose steady states have mixed-state symmetry-protected-topological order in the bulk, with corresponding edge theories characterized by nontrivial anomalies.

quant-ph↗
arXiv2024

Fewer Truncations Improve Language Modeling

In large language model training, input documents are typically concatenated together and then split into sequences of equal length to avoid padding tokens. Despite its efficiency, the concatenation approach compromises data integrity -- it inevitably breaks many documents into incomplete pieces, leading to excessive truncations that hinder the model from learning to compose logically coherent and factually consistent content that is grounded on the complete context. To address the issue, we propose Best-fit Packing, a scalable and efficient method that packs documents into training sequences through length-aware combinatorial optimization. Our method completely eliminates unnecessary truncations while retaining the same training efficiency as concatenation. Empirical results from both text and code pre-training show that our method achieves superior performance (e.g., relatively +4.7% on reading comprehension; +16.8% in context following; and +9.2% on program synthesis), and reduces closed-domain hallucination effectively by up to 58.3%.

cs.CL↗
arXiv2024

A proposed PKU-Muon experiment for muon tomography and dark matter search

We propose here a set of new methods to directly detect light mass dark matter through its scattering with abundant atmospheric muons or accelerator beams. Firstly, we plan to use the free cosmic-ray muons interacting with dark matter in a volume surrounded by tracking detectors, to trace possible interaction between dark matter and muons. Secondly, we will interface our device with domestic or international muon beams. Due to much larger muon intensity and focused beam, we anticipate the detector can be made further compact and the resulting sensitivity on dark matter searches will be improved. Furthermore, we will measure precisely directional distributions of cosmic-ray muons, either at mountain or sea level, and the differences may reveal possible information of dark matter distributed near the earth. Specifically, our methods can have advantages over `exotic' dark matters which are either muon-philic or slowed down due to some mechanism, and sensitivity on dark matter and muon scattering cross section can reach as low as microbarn level.

hep-ex↗
arXiv2024

CIFAR-10-Warehouse: Broad and More Realistic Testbeds in Model Generalization Analysis

Analyzing model performance in various unseen environments is a critical research problem in the machine learning community. To study this problem, it is important to construct a testbed with out-of-distribution test sets that have broad coverage of environmental discrepancies. However, existing testbeds typically either have a small number of domains or are synthesized by image corruptions, hindering algorithm design that demonstrates real-world effectiveness. In this paper, we introduce CIFAR-10-Warehouse, consisting of 180 datasets collected by prompting image search engines and diffusion models in various ways. Generally sized between 300 and 8,000 images, the datasets contain natural images, cartoons, certain colors, or objects that do not naturally appear. With CIFAR-10-W, we aim to enhance the evaluation and deepen the understanding of two generalization tasks: domain generalization and model accuracy prediction in various out-of-distribution environments. We conduct extensive benchmarking and comparison experiments and show that CIFAR-10-W offers new and interesting insights inherent to these tasks. We also discuss other fields that would benefit from CIFAR-10-W.

cs.CV↗