arXiv subjects

Li Ju

Publications and source records attributed to Li Ju.

At least 19 recordsLinked to original sources

arXiv2026

NovGauge: A Fine-Grained Benchmark for Diagnosing LLMs' Capability in Paper Novelty Assessment

Large language models (LLMs) are increasingly used in peer review at major AI conferences, yet novelty remains a persistent weak point. Existing benchmarks assess novelty as a single holistic score, making it difficult to diagnose which dimension a model misjudges or whether its evidence is faithful. We present NovGauge, a human-anchored benchmark for fine-grained novelty assessment diagnosis. The benchmark contains 619 paper pairs and 50 multi-paper sets, drawn from two expert sources: ICLR reviewer overlap claims and survey co-citations. Instances are independently labeled along three dimensions: task, problem, and method, capturing application goals, technical challenges, and solution approaches. We propose a cascading diagnostic pipeline that verifies per-dimension correctness, evidence grounding, and logical support. Evaluation of 18 LLMs shows hallucination rates ranging from 0% to 39% across dimensions, and among non-hallucinated correct-positive judgments, over 70% cite evidence fails to logically support the stated reason. The best-performing model, GPT-5.5, achieves 43-72% Verified F1 across dimensions, while most models retain less than half of their raw F1 after faithfulness verification. These results suggest that current LLMs remain far from reliable scientific novelty assessment, particularly when correctness is conditioned on faithful evidence grounding.

cs.AI↗

arXiv2026

First Demonstration of Optical Feedback Control to Parametric Instability at Advanced LIGO

Increasing the circulating power in gravitational-wave detectors to the megawatt level is essential for future sensitivity improvement, but this is critically limited by optomechanical parametric instabilities. Current mitigation strategies are projected to be inadequate against instabilities when circulating power reaches a megawatt. Optical feedback offers a novel independent paradigm to mitigate parametric instability. In this Letter, we report the first demonstration of optical feedback control in a full-scale gravitational wave detector. We successfully suppressed an unstable mode at 10.428 kHz, reducing the parametric gain from R = 2 to R < 0.02. This work validates optical feedback control as an effective mitigation scheme for kilometre-scale interferometric gravitational-wave detectors, providing an effective strategy to allow detectors to reach the megawatt level.

gr-qc↗

arXiv2026

A Novel Arm-Length Stabilization Scheme for Gravitational-Wave Detectors with AlGaAs/GaAs Coated Mirrors

The arm length stabilisation system is employed in gravitational-wave detectors to reduce the velocity of the mirrors such that the arm cavities can be brought onto resonance in a controlled manner required to attain the detector operating point. For future upgrades of current gravitational wave detectors such as A#, which will incorporate AlGaAs/GaAs coatings, the current frequency-doubled arm length stabilisation system is unsuitable due to excessive absorption of the frequency-doubled 532nm beam by the AlGaAs/GaAs coating. We propose a novel multi-wavelength arm length stabilisation scheme that uses both frequency-doubled and frequency-tripled beams. The 1596nm auxiliary locking beam is outside the absorption bands of AlGaAs/GaAs coating. It is frequencytripled to 532nm and phase-locked with the 1064nm science laser through its second harmonic at 532 nm. In a tabletop setup, we experimentally demonstrated the stable cavity detuning and robust cavity locking transition by controlling the 1596nm laser and 1064nm laser phase locked loop. This demonstration confirmed that the proposed novel arm length stabilisation scheme is compatible with future upgrades or third-generation gravitational wave detectors that use AlGaAs/GaAs-coated test masses.

physics.optics↗

arXiv2026

GeoFlowVLM: Geometry-Aware Joint Uncertainty for Frozen Vision-Language Embedding

Standard dual-encoder vision-language models that map images and text to deterministic points on a shared unit hypersphere through $\ell_2$ normalization typically expose neither \emph{aleatoric} uncertainty (cross-modal ambiguity) nor \emph{epistemic} uncertainty (lack of training-distribution support). Existing post-hoc methods either recover at most one of the two uncertainty components, or ignore the hyperspherical geometry of these models' embeddings. We propose \textbf{GeoFlowVLM} as a post-hoc adapter that learns the joint distribution of paired $\ell_2$-normalised dual-encoder VLM embeddings on the product hypersphere $\mathbb{S}^{d-1} \times \mathbb{S}^{d-1}$ via Riemannian flow matching with a single masked velocity field. A consistency result shows that, in the population limit, the trained network exposes the joint flow and both cross-modal conditional flows as valid Riemannian flow-matching velocity fields on their respective domains. We derive two quantities from this single model: a conditional retrieval entropy that quantifies aleatoric ambiguity with a decision-theoretic interpretation via a Fano-type bound, and a marginal-typicality epistemic score justified by an exact chain-rule decomposition of the joint NLL. This decomposition isolates a cross-modal pointwise-mutual-information term that is structurally discriminative rather than epistemic, and is empirically the only consistently uninformative standalone component. Empirically, the entropy tracks Recall@1 with near-ideal monotonic calibration across three retrieval benchmarks in both directions, and the marginal-typicality sum yields consistently calibrated selective accuracy across four zero-shot classification benchmarks.

cs.LG↗

arXiv2026

Faithfulness-QA: A Counterfactual Entity Substitution Dataset for Training Context-Faithful RAG Models

Retrieval-Augmented Generation (RAG) models frequently produce answers grounded in parametric memory rather than the retrieved context, undermining the core promise of retrieval augmentation. A fundamental obstacle to fixing this unfaithfulness is the lack of training data that explicitly requires models to prefer context over internal knowledge. We introduce Faithfulness-QA, a large-scale dataset of 99,094 samples constructed through counterfactual entity substitution. Starting from two established extractive QA benchmarks--SQuAD and TriviaQA--we automatically identify answer-bearing named entities in each context, replace them with type-consistent alternatives drawn from a curated bank of 76,953 entities, and thereby manufacture controlled knowledge conflicts between context and parametric memory. Rigorous quality filtering ensures 100% pass rates across four automated checks on random 200-sample audits. We release the full dataset, the construction pipeline, and a typed entity bank covering eight named entity categories. Faithfulness-QA is designed as a training resource for attention-based faithfulness objectives and as an evaluation benchmark for measuring context-grounding behavior in RAG systems. Data and code are available at https://github.com/qzhangFDU/faithfulness-qa-dataset.

cs.CL↗

arXiv2026

OneFlowSBI: One Model, Many Queries for Simulation-Based Inference

We introduce \textit{OneFlowSBI}, a unified framework for simulation-based inference that learns a single flow-matching generative model over the joint distribution of parameters and observations. Leveraging a query-aware masking distribution during training, the same model supports multiple inference tasks, including posterior sampling, likelihood estimation, and arbitrary conditional distributions, without task-specific retraining. We evaluate \textit{OneFlowSBI} on ten benchmark inference problems and two high-dimensional real-world inverse problems across multiple simulation budgets. \textit{OneFlowSBI} is shown to deliver competitive performance against state-of-the-art generalized inference solvers and specialized posterior estimators, while enabling efficient sampling with few ODE integration steps and remaining robust under noisy and partially observed data.

stat.ML↗

arXiv2026

Epistemic Uncertainty Quantification for Pre-trained VLMs via Riemannian Flow Matching

Vision-Language Models (VLMs) are typically deterministic in nature and lack intrinsic mechanisms to quantify epistemic uncertainty, which reflects the model's lack of knowledge or ignorance of its own representations. We theoretically motivate negative log-density of an embedding as a proxy for the epistemic uncertainty, where low-density regions signify model ignorance. The proposed method REPVLM computes the probability density on the hyperspherical manifold of the VLM embeddings using Riemannian Flow Matching. We empirically demonstrate that REPVLM achieves near-perfect correlation between uncertainty and prediction error, significantly outperforming existing baselines. Beyond classification, we also demonstrate that the model also provides a scalable metric for out-of-distribution detection and automated data curation.

cs.LG↗

arXiv2026

OpenNovelty: An LLM-powered Agentic System for Verifiable Scholarly Novelty Assessment

Evaluating novelty is critical yet challenging in peer review, as reviewers must assess submissions against a vast, rapidly evolving literature. This report presents OpenNovelty, an LLM-powered agentic system for transparent, evidence-based novelty analysis. The system operates through four phases: (1) extracting the core task and contribution claims to generate retrieval queries; (2) retrieving relevant prior work based on extracted queries via semantic search engine; (3) constructing a hierarchical taxonomy of core-task-related work and performing contribution-level full-text comparisons against each contribution; and (4) synthesizing all analyses into a structured novelty report with explicit citations and evidence snippets. Unlike naive LLM-based approaches, \textsc{OpenNovelty} grounds all assessments in retrieved real papers, ensuring verifiable judgments. We deploy our system on 500+ ICLR 2026 submissions with all reports publicly available on our website, and preliminary analysis suggests it can identify relevant prior work, including closely related papers that authors may overlook. OpenNovelty aims to empower the research community with a scalable tool that promotes fair, consistent, and evidence-backed peer review.

cs.IR↗

arXiv2025

WisPaper: Your AI Scholar Search Engine

We present \textsc{WisPaper}, an end-to-end agent system that transforms how researchers discover, organize, and track academic literature. The system addresses two fundamental challenges. (1)~\textit{Semantic search limitations}: existing academic search engines match keywords but cannot verify whether papers truly address complex research questions; and (2)~\textit{Workflow fragmentation}: researchers must manually stitch together separate tools for discovery, organization, and monitoring. \textsc{WisPaper} tackles these through three integrated modules. \textbf{Scholar Search} combines rapid keyword retrieval with \textit{Deep Search}, in which an agentic model, \textsc{WisModel}, validates candidate papers against user queries through structured reasoning. Discovered papers flow seamlessly into \textbf{Library} with one click, where systematic organization progressively builds a user profile that sharpens the recommendations of \textbf{AI Feeds}, which continuously surfaces relevant new publications and in turn guides subsequent exploration, closing the loop from discovery to long-term awareness. On TaxoBench, \textsc{WisPaper} achieves 22.26\% recall, surpassing the O3 baseline (20.92\%). Furthermore, \textsc{WisModel} attains 93.70\% validation accuracy, effectively mitigating retrieval hallucinations.

cs.IR↗

arXiv2025

Seismic Isolation of Optical Tables Using Piezo Actuators

Seismic isolation is crucial for gravitational wave detectors as it minimizes ground vibrations, enabling the detection of faint gravitational wave signals. An active seismic isolation platform for precision measurement experiments is described. The table features piezo actuation along five degrees of freedom: three translational actuations and two tip-tilt degrees of freedom along the horizontal axes. It is stiff in rotation about the vertical axes. A seismometer is used to sense table motion. Piezo actuators are used to suppress seismic noise with feedback control bandwidth of 0.3 to 3 Hz. Suppression levels ranging from 21 to 36 dB of seismic noise within the frequency range of 0.5 to 1.3 Hz are demonstrated, as measured by a witness seismometer on the table, with the suppression direction along the axis of the longitudinal translation of the suspended mirror on the table. The suppression results in 1 $\mathrm{\mathrm{nm/\sqrt{Hz}}}$ residual horizontal motion at 1 Hz. Limitations such as tilt-to-translation coupling that prevent actuation over the desired range of 0.03 to 3 Hz are discussed.

astro-ph.IM↗

arXiv2025

Exploiting the Asymmetric Uncertainty Structure of Pre-trained VLMs on the Unit Hypersphere

Vision-language models (VLMs) as foundation models have significantly enhanced performance across a wide range of visual and textual tasks, without requiring large-scale training from scratch for downstream tasks. However, these deterministic VLMs fail to capture the inherent ambiguity and uncertainty in natural language and visual data. Recent probabilistic post-hoc adaptation methods address this by mapping deterministic embeddings onto probability distributions; however, existing approaches do not account for the asymmetric uncertainty structure of the modalities, and the constraint that meaningful deterministic embeddings reside on a unit hypersphere, potentially leading to suboptimal performance. In this paper, we address the asymmetric uncertainty structure inherent in textual and visual data, and propose AsymVLM to build probabilistic embeddings from pre-trained VLMs on the unit hypersphere, enabling uncertainty quantification. We validate the effectiveness of the probabilistic embeddings on established benchmarks, and present comprehensive ablation studies demonstrating the inherent nature of asymmetry in the uncertainty structure of textual and visual data.

cs.LG↗

arXiv2025

PARIC: Probabilistic Attention Regularization for Language Guided Image Classification from Pre-trained Vison Language Models

Language-guided attention frameworks have significantly enhanced both interpretability and performance in image classification; however, the reliance on deterministic embeddings from pre-trained vision-language foundation models to generate reference attention maps frequently overlooks the intrinsic multivaluedness and ill-posed characteristics of cross-modal mappings. To address these limitations, we introduce PARIC, a probabilistic framework for guiding visual attention via language specifications. Our approach enables pre-trained vision-language models to generate probabilistic reference attention maps, which align textual and visual modalities more effectively while incorporating uncertainty estimates, as compared to their deterministic counterparts. Experiments on benchmark test problems demonstrate that PARIC enhances prediction accuracy, mitigates bias, ensures consistent predictions, and improves robustness across various datasets.

cs.CV↗

arXiv2025

GraphBridge: Towards Arbitrary Transfer Learning in GNNs

Graph neural networks (GNNs) are conventionally trained on a per-domain, per-task basis. It creates a significant barrier in transferring the acquired knowledge to different, heterogeneous data setups. This paper introduces GraphBridge, a novel framework to enable knowledge transfer across disparate tasks and domains in GNNs, circumventing the need for modifications to task configurations or graph structures. Specifically, GraphBridge allows for the augmentation of any pre-trained GNN with prediction heads and a bridging network that connects the input to the output layer. This architecture not only preserves the intrinsic knowledge of the original model but also supports outputs of arbitrary dimensions. To mitigate the negative transfer problem, GraphBridge merges the source model with a concurrently trained model, thereby reducing the source bias when applied to the target domain. Our method is thoroughly evaluated across diverse transfer learning scenarios, including Graph2Graph, Node2Node, Graph2Node, and graph2point-cloud. Empirical validation, conducted over 16 datasets representative of these scenarios, confirms the framework's capacity for task- and domain-agnostic transfer learning within graph-like data, marking a significant advancement in the field of GNNs. Code is available at https://github.com/jujulili888/GraphBridge.

cs.LG↗

arXiv2025

$\texttt{InfoHier}$: Hierarchical Information Extraction via Encoding and Embedding

Analyzing large-scale datasets, especially involving complex and high-dimensional data like images, is particularly challenging. While self-supervised learning (SSL) has proven effective for learning representations from unlabelled data, it typically focuses on flat, non-hierarchical structures, missing the multi-level relationships present in many real-world datasets. Hierarchical clustering (HC) can uncover these relationships by organizing data into a tree-like structure, but it often relies on rigid similarity metrics that struggle to capture the complexity of diverse data types. To address these we envision $\texttt{InfoHier}$, a framework that combines SSL with HC to jointly learn robust latent representations and hierarchical structures. This approach leverages SSL to provide adaptive representations, enhancing HC's ability to capture complex patterns. Simultaneously, it integrates HC loss to refine SSL training, resulting in representations that are more attuned to the underlying information hierarchy. $\texttt{InfoHier}$ has the potential to improve the expressiveness and performance of both clustering and representation learning, offering significant benefits for data analysis, management, and information retrieval.

cs.IR↗

arXiv2024

Demonstration of optical spring in an un-detuned cavity containing an optical parametric amplifier

Here we demonstrate the capacity to manipulate the optical spring (OS) effect by employing an optical parametric amplifier (OPA) within an optical cavity. We observed more than a factor of 2 increase in the OS frequency shift with the OPA. We also showed for the first time that the OS can be tuned by solely adjusting the OPA phase and showing an un-detuned cavity exhibiting an optical spring. The method can be applied to gravitational wave detectors in the signal recycling configuration to realize narrow bandwidth high sensitivity. The OS can be tuned to align the detector peak sensitivity frequency to known frequency continuous gravitational wave signals, dynamically tuned to track the gravitational wave signal from merging compact binaries or tuned to search for the post-merger signal of known binary coalescence.

physics.optics↗

arXiv2024

SGV3D:Towards Scenario Generalization for Vision-based Roadside 3D Object Detection

Roadside perception can greatly increase the safety of autonomous vehicles by extending their perception ability beyond the visual range and addressing blind spots. However, current state-of-the-art vision-based roadside detection methods possess high accuracy on labeled scenes but have inferior performance on new scenes. This is because roadside cameras remain stationary after installation and can only collect data from a single scene, resulting in the algorithm overfitting these roadside backgrounds and camera poses. To address this issue, in this paper, we propose an innovative Scenario Generalization Framework for Vision-based Roadside 3D Object Detection, dubbed SGV3D. Specifically, we employ a Background-suppressed Module (BSM) to mitigate background overfitting in vision-centric pipelines by attenuating background features during the 2D to bird's-eye-view projection. Furthermore, by introducing the Semi-supervised Data Generation Pipeline (SSDG) using unlabeled images from new scenes, diverse instance foregrounds with varying camera poses are generated, addressing the risk of overfitting specific camera poses. We evaluate our method on two large-scale roadside benchmarks. Our method surpasses all previous methods by a significant margin in new scenes, including +42.57% for vehicle, +5.87% for pedestrian, and +14.89% for cyclist compared to BEVHeight on the DAIR-V2X-I heterologous benchmark. On the larger-scale Rope3D heterologous benchmark, we achieve notable gains of 14.48% for car and 12.41% for large vehicle. We aspire to contribute insights on the exploration of roadside perception techniques, emphasizing their capability for scenario generalization. The code will be available at https://github.com/yanglei18/SGV3D

cs.CV↗

arXiv2023

Single and coupled cavity mode sensing schemes using a diagnostic field

Precise optical mode matching is of critical importance in experiments using squeezed-vacuum states. Automatic spatial-mode matching schemes have the potential to reduce losses and improve loss stability. However, in quantum-enhanced coupled-cavity experiments, such as gravitational-wave detectors, one must also ensure that the sub-cavities are also mode matched. We propose a new mode sensing scheme, which works for simple and coupled cavities. The scheme requires no moving parts, nor tuning of Gouy phases. Instead a diagnostic field tuned to the HG20/LG10 mode frequency is used. The error signals are derived to be proportional to the difference in waist position, and difference in Rayleigh ranges, between the sub-cavity eigenmodes. The two error signals are separable by 90 degrees of demodulation phase. We demonstrate reasonable error signals for a simplified Einstein Telescope optical design. This work will facilitate routine use of extremely high levels of squeezing in current and future gravitational-wave detectors.

physics.optics↗

arXiv2023

Developing and implementing an Einsteinian science curriculum from Years 3 to 10 : Part A Concepts, rationale and learning outcomes

There has been a growing realisation that school science curricula do not adequately reflect the revolutionary changes in our scientific understanding of the 20th century. This discrepancy between current school education and our modern scientific understanding has led to calls for the modernisation of the science curriculum. Although there have been attempts to introduce topics of Einsteinian physics (i.e., quantum physics and relativity) to school education, often at the secondary level, we still lack a seamless curriculum in which modern science concepts are gradually introduced in primary and middle schools. Guided by the Model of Educational Reconstruction and following a mixed-methods research design, the Einstein-First project aims to address this gap. Einstein-First has developed and implemented an Einsteinian curriculum from Years 3 to 10 (students aged 7- 16) that resolves the disconnect between science in schools and the modern world. This paper presents the concepts, rationale, and learning outcomes of the curriculum implementation in six Australian schools with 315 students across Years 3 to 10. Our findings lay the foundation for informed curriculum development towards a school education that can enhance students' understanding and appreciation of the fundamental concepts of modern science and its impact on our society.

physics.ed-ph↗