arXiv subjects

Yiqing Liu

Publications and source records attributed to Yiqing Liu.

At least 19 recordsLinked to original sources

arXiv2026

Shared Prefixes, Better Credit: Adaptive Routing for Multi-Agent Reasoning

Multi-agent reasoning (MAR) improves reasoning reliability through iterative solution exchange and refinement. Existing adaptive MAR methods typically learn routing decisions from query-level labels or trajectory-level returns, but such coarse supervision cannot accurately estimate the state-conditioned utility of individual operators in multi-step collaboration. We propose TreeCredit, a shared-prefix credit assignment framework for efficient adaptive MAR. Its core insight is to estimate operator utility through state-matched downstream comparisons, rather than directly attributing trajectory-level outcomes to preceding decisions. TreeCredit constructs shared-prefix collaboration trees by expanding candidate operators from the same intermediate state and assigns each state--operator pair a correctness-prioritized suffix credit based on the terminal correctness and cumulative additional cost of its complete continuation. These structured credits are converted into state-local operator preferences to train a lightweight pairwise state router, which dynamically selects the next admissible operator during inference. Experiments on six reasoning benchmarks show that TreeCredit modestly improves accuracy while substantially reducing inference cost, achieving a better accuracy--cost trade-off than representative MAR methods.

cs.AI↗

arXiv2026

EMS: Multi-Agent Voting via Efficient Majority-then-Stopping

Majority voting is the standard for aggregating multi-agent responses into a final decision. However, traditional methods typically require all agents to complete their reasoning before aggregation begins, leading to significant computational overhead, as many responses become redundant once a majority consensus is achieved. In this work, we formulate efficient multi-agent voting as a reliability-aware agent scheduling problem and propose Efficient Majority-then-Stopping (EMS) to improve reasoning efficiency. EMS first estimates a Task-Conditioned Reliability Ordering (TCRO) for each agent by retrieving its historical consensus evidence on semantically similar queries, and then invoking agents in descending reliability order. Next, Adaptive Incremental Voting (AIV) terminates the process once the current leading answer cannot be overturned by any possible votes from the remaining agents, and returns this answer. Finally, Reliability History Updating (RHU) updates only the invoked agents according to their consensus with the final decision. Extensive evaluations across five benchmarks show that EMS preserves the accuracy of Majority Voting while reducing the average number of invoked agents by 35% and token consumption by 44%, respectively. The code is available at https://github.com/fuyu66/EMS.

cs.AI↗

arXiv2026

HCP-MAD:Heterogeneous Consensus-Progressive Reasoning for Efficient Multi-Agent Debate

Multi-Agent Debate (MAD) is a collaborative framework in which multiple agents iteratively refine solutions through the generation of reasoning and alternating critique cycles. Current work primarily optimizes intra-round topologies and inter-round interactions separately, limiting the adaptation of token costs to task complexity. This work introduces Heterogeneous Consensus-Progressive Reasoning for Efficient Multi-Agent Debate (HCP-MAD), leveraging consensus as a dynamic signal to facilitate progressive reasoning. The core motivation is that a majority of straightforward tasks can be effectively resolved via lightweight pair-agent debates, while complex tasks require expanded collaboration. Firstly, Heterogeneous Consensus Verification conducts rapid consensus verification using a pair of heterogeneous agents for early stopping. Next, Heterogeneous Pair-Agent Debate applies an adaptive stopping criterion to terminate mutual critique of reasoning traces. Finally, the unresolved tasks are addressed through Escalated Collective Voting by aggregating diverse perspectives from additional agents. Experiments across six benchmarks show that HCP-MAD enhances accuracy while substantially reducing token costs. Code is https://github.com/fuyu66/HCP-MAD.

cs.MA↗

arXiv2026

GloPath: An Entity-Centric Foundation Model for Glomerular Lesion Assessment and Clinicopathological Insights

Glomerular pathology is central to the diagnosis and prognosis of renal diseases, yet the heterogeneity of glomerular morphology and fine-grained lesion patterns remain challenging for current AI approaches. We present GloPath, an entity-centric foundation model trained on over one million glomeruli extracted from 14,049 renal biopsy specimens using multi-scale and multi-view self-supervised learning. GloPath addresses two major challenges in nephropathology: glomerular lesion assessment and clinicopathological insights discovery. For lesion assessment, GloPath was benchmarked across three independent cohorts on 52 tasks, including lesion recognition, grading, few-shot classification, and cross-modality diagnosis-outperforming state-of-the-art methods in 42 tasks (80.8%). In the large-scale real-world study, it achieved an ROC-AUC of 91.51% for lesion recognition, demonstrating strong robustness in routine clinical settings. For clinicopathological insights, GloPath systematically revealed statistically significant associations between glomerular morphological parameters and clinical indicators across 224 morphology-clinical variable pairs, demonstrating its capacity to connect tissue-level pathology with patient-level outcomes. Together, these results position GloPath as a scalable and interpretable platform for glomerular lesion assessment and clinicopathological discovery, representing a step toward clinically translatable AI in renal pathology.

cs.CV↗

arXiv2026

Dynamic Dispersion Accumulation in Fiber Loops for Realizing Record-High Frequency Resolution or Ultra-Low Signal Sampling Rate in Dispersion-Based Photonics-Assisted Wideband Microwave Measurement Systems

Dispersion-based photonics-assisted microwave measurement systems provide immense potential for real-time analysis of wideband and dynamic signals. However, they face two critical challenges: a difficulty in achieving high frequency resolution over a wideband analysis bandwidth, and a reliance on large-bandwidth-and-high-sampling-rate oscilloscopes to capture the resulting ultra-narrow pulses. We introduce a dynamic dispersion accumulation technique to overcome these limitations. By circulating the optical signal in fiber loops containing a dispersion-compensating fiber, we achieve a high accumulated dispersion of -215700 ps/nm. This high dispersion relaxes the required chirp rate of the chirped optical signal, enabling two distinct advantages: When the analysis bandwidth is fixed, a lower chirp rate enables a longer temporal period, yielding a record-high frequency resolution of 27.9 MHz; When the temporal period is fixed, a lower chirp rate enables a smaller bandwidth, generating a wider pulse and thus relaxing pulse sampling requirements at the expense of analysis bandwidth. This sacrifice in analysis bandwidth can be compensated by a duty-cycle-enabling technique, which holds the potential to extend the analysis bandwidth beyond 100 GHz. This work breaks the performance and hardware limitations in dispersion-based systems, paving the way for high frequency resolution, wideband microwave measurement systems that are both real-time and cost-effective.

physics.optics↗

arXiv2025

A benchmark dataset for evaluating Syndrome Differentiation and Treatment in large language models

The emergence of Large Language Models (LLMs) within the Traditional Chinese Medicine (TCM) domain presents an urgent need to assess their clinical application capabilities. However, such evaluations are challenged by the individualized, holistic, and diverse nature of TCM's "Syndrome Differentiation and Treatment" (SDT). Existing benchmarks are confined to knowledge-based question-answering or the accuracy of syndrome differentiation, often neglecting assessment of treatment decision-making. Here, we propose a comprehensive, clinical case-based benchmark spearheaded by TCM experts, and a specialized reward model employed to quantify prescription-syndrome congruence. Data annotation follows a rigorous pipeline. This benchmark, designated TCM-BEST4SDT, encompasses four tasks, including TCM Basic Knowledge, Medical Ethics, LLM Content Safety, and SDT. The evaluation framework integrates three mechanisms, namely selected-response evaluation, judge model evaluation, and reward model evaluation. The effectiveness of TCM-BEST4SDT was corroborated through experiments on 15 mainstream LLMs, spanning both general and TCM domains. To foster the development of intelligent TCM research, TCM-BEST4SDT is now publicly available.

cs.CL↗

arXiv2025

From Generic to Specialized: A Subspecialty Diagnostic System Powered by Self-Supervised Learning for Cervical Histopathology

Cervical cancer remains a major malignancy, necessitating extensive and complex histopathological assessments and comprehensive support tools. Although deep learning shows promise, these models still lack accuracy and generalizability. General foundation models offer a broader reach but remain limited in capturing subspecialty-specific features and task adaptability. We introduce the Cervical Subspecialty Pathology (CerS-Path) diagnostic system, developed through two synergistic pretraining stages: self-supervised learning on approximately 190 million tissue patches from 140,000 slides to build a cervical-specific feature extractor, and multimodal enhancement with 2.5 million image-text pairs, followed by integration with multiple downstream diagnostic functions. Supporting eight diagnostic functions, including rare cancer classification and multimodal Q&A, CerS-Path surpasses prior foundation models in scope and clinical applicability. Comprehensive evaluations demonstrate a significant advance in cervical pathology, with prospective testing on 3,173 cases across five centers maintaining 99.38% screening sensitivity and excellent generalizability, highlighting its potential for subspecialty diagnostic translation and cervical cancer screening.

cs.CV↗

arXiv2025

FECT: Classification of Breast Cancer Pathological Images Based on Fusion Features

Breast cancer is one of the most common cancers among women globally, with early diagnosis and precise classification being crucial. With the advancement of deep learning and computer vision, the automatic classification of breast tissue pathological images has emerged as a research focus. Existing methods typically rely on singular cell or tissue features and lack design considerations for morphological characteristics of challenging-to-classify categories, resulting in suboptimal classification performance. To address these problems, we proposes a novel breast cancer tissue classification model that Fused features of Edges, Cells, and Tissues (FECT), employing the ResMTUNet and an attention-based aggregator to extract and aggregate these features. Extensive testing on the BRACS dataset demonstrates that our model surpasses current advanced methods in terms of classification accuracy and F1 scores. Moreover, due to its feature fusion that aligns with the diagnostic approach of pathologists, our model exhibits interpretability and holds promise for significant roles in future clinical applications.

eess.IV↗

arXiv2024

Microwave photonic frequency measurement and time-frequency analysis: Unlocking bandwidths over hundreds of GHz with a 10-nanosecond temporal resolution

Fast and broadband spectrum sensing is an essential component in cognitive radio systems, intelligent transportation systems, electronic warfare systems, etc. However, traditional electronic-based solutions have a trade-off among the analysis bandwidth, temporal resolution, and real-time performance. In comparison, microwave photonic solutions can overcome the trade-off at the cost of frequency accuracy and resolution. Nevertheless, the reported microwave photonic solutions suffer from a very poor frequency resolution and impose extremely high requirements on hardware when the analysis bandwidth is close to or greater than 100 GHz. Here, we show a microwave photonic frequency measurement and time-frequency analysis method, which is implemented by dispersion-based frequency-to-time mapping and assisted by a specially designed V-shape linearly frequency-modulated signal and a duty-cycle-enabling technique. Compared with the reported microwave photonic solutions, the hardware requirements are greatly reduced when achieving similar performance conditions. Using a total dispersion of -6817 ps/nm and a V-shape linearly frequency-modulated signal with a bandwidth of 31.6 GHz and a duty cycle of 1/4, we achieve an ambiguity-free analysis bandwidth of 252.8 GHz, a corresponding temporal resolution of 13.75 ns and a frequency resolution of 1.1 GHz. The temporal resolution can be improved to 6.875 ns when the duty cycle is changed to 1/2, while the analysis bandwidth in this case is 126.4 GHz.

physics.optics↗

arXiv2024

Shapley Values-enabled Progressive Pseudo Bag Augmentation for Whole Slide Image Classification

In computational pathology, whole-slide image (WSI) classification presents a formidable challenge due to its gigapixel resolution and limited fine-grained annotations. Multiple-instance learning (MIL) offers a weakly supervised solution, yet refining instance-level information from bag-level labels remains challenging. While most of the conventional MIL methods use attention scores to estimate instance importance scores (IIS) which contribute to the prediction of the slide labels, these often lead to skewed attention distributions and inaccuracies in identifying crucial instances. To address these issues, we propose a new approach inspired by cooperative game theory: employing Shapley values to assess each instance's contribution, thereby improving IIS estimation. The computation of the Shapley value is then accelerated using attention, meanwhile retaining the enhanced instance identification and prioritization. We further introduce a framework for the progressive assignment of pseudo bags based on estimated IIS, encouraging more balanced attention distributions in MIL models. Our extensive experiments on CAMELYON-16, BRACS, TCGA-LUNG, and TCGA-BRCA datasets show our method's superiority over existing state-of-the-art approaches, offering enhanced interpretability and class-wise insights. Our source code is available at https://github.com/RenaoYan/PMIL.

cs.CV↗

arXiv2024

Examining Differential Item Functioning (DIF) in Self-Reported Health Survey Data: Via Multilevel Modeling

Few health-related constructs or measures have received a critical evaluation in terms of measurement equivalence, such as self-reported health survey data. Differential item functioning (DIF) analysis is crucial for evaluating measurement equivalence in self-reported health surveys, which are often hierarchical in structure. Traditional single-level DIF methods in this case fall short, making multilevel models a better alternative. We highlight the benefits of multilevel modeling for DIF analysis, when applying a health survey data set to multilevel binary logistic regression (for analyzing binary response data) and multilevel multinominal logistic regression (for analyzing polytomous response data), and comparing them with their single-level counterparts. Our findings show that multilevel models fit better and explain more variance than single-level models. This article is expected to raise awareness of multilevel modeling and help healthcare researchers and practitioners understand the use of multilevel modeling for DIF analysis.

stat.AP↗

arXiv2024

Physical encryption and decryption for secure data transmission in optical networks leveraging the temporal Talbot effect and microwave photonics

A novel microwave photonic scheme for secure data transmission in optical networks is proposed. The security of the scheme is guaranteed by physical encryption and decryption via the temporal Talbot effect in dispersive mediums. First, the original data is randomized in the digital domain by performing an exclusive OR operation using a random matrix. Subsequently, a time-varying multi-tone electrical signal, which represents the randomized data matrix, is modulated onto an optical carrier. The optical signal after modulation is then phase-modulated by a temporal Talbot array illuminator (TAI) signal, and the optical signal after discrete quadratic phase modulation will lose its original appearance in the frequency domain and be further dispersed in the first dispersive medium. Due to the dispersion that does not match the TAI signal exactly, the waveform after the first dispersive medium is a noise-like signal. Hence, the physical encryption of the original data is successfully achieved. As the optical signal passes a second dispersive medium that makes the total dispersion match the TAI signal, the temporal waveform of the noise-like signal after photodetection is transformed into pulses. "1" and "0" in the randomized data matrix are represented through the presence and absence of pulses, and the physical decryption is achieved. By further processing the recovered data matrix using the random matrix, the original data can be recovered. The physical layer security of the proposed scheme and its fiber transmission capability are demonstrated. 8-Gbit/s data is transmitted, encrypted, and decrypted using two dispersive mediums and an optical fiber of 10 to 200 km, and error-free transmission is achieved. Many factors that affect the encryption, decryption, and transmission performance of the system have been analyzed.

physics.optics↗

arXiv2020

Single-Sideband Time-Modulated Phased Array With 2-bit Phased Shifters

A novel single-sideband (SSB) time-modulated technique with 2-bit phase shifters is proposed. The timemodulated module is implemented by adding periodic phase modulation to 2-bit phase shifters, which is simpler without performance loss compared to existing SSB time-modulated method. During one modulation period, four phase states (0, π/2, π, 3π/2) of 2-bit phase shifters are switched in sequence. After the modulation, the SSB time modulation is realized and the main power is distributed to the first harmonic component. The feasibility of the proposed method is verified by experiments. The undesired harmonics are efficiently suppressed. Meanwhile, 80° beam scanning range are realized through the proposed module.

eess.SP↗

arXiv2020

SDSS-IV MaNGA: The [$α$/Fe] of Early-Type Galaxies

The mean stellar alpha-to-iron abundance ratio ([$α$/Fe]) of a galaxy is an indicator of galactic star formation timescale. It is important for understanding the star formation history of early-type galaxies (ETGs) as their star formation processes have basically stopped. Using the model templates which are made by Vazdekis et al., we apply the pPXF based spectral fitting method to estimate the [$α$/Fe] of 196 high signal-to-noise ratio ETGs from the MaNGA survey. The velocity dispersions within 1R$_e$ ($σ_{e}$) range from 27 to 270 km/s. We find a flat relation between the mean [$α$/Fe] within the 1R$_e^{maj}$ ellipses and log($σ_{e}$), even if limiting to the massive sample with log($σ_{e}$/km s$^{-1}$)$>$1.9. However, the relation becomes positive after we exclude the Mg$_1$ feature in our fits, which agrees with the results from the previous work with other stellar population models, albeit with relatively large scatter. It indicates that the spectral fits with Vazdekis models could give basically the consistent predictions of [$α$/Fe] with previous studies when the Mg$_b$ index is used, but do not work well at the Mg$_1$ band when their $α$-enhanced version is employed in the metal-rich regime. We suggest avoiding this rather wide index, which covers 471Å, as it might suffer from other effects such as flux-calibration issues. For reference, we also measure the stellar population radial gradients within 1R$_e^{maj}$ ellipses. Due to the low resolution of age estimations for old objects and the Mg$_1$ issue, the uncertainties of these gradients cannot be neglected.

astro-ph.GA↗

arXiv2019

The ACS Fornax Cluster Survey. III. Globular Cluster Specific Frequencies of Early-Type Galaxies

The globular cluster (GC) specific frequency ($S_N$), defined as the number of GCs per unit galactic luminosity, represents the efficiency of GC formation (and survival) compared to field stars. Despite the naive expectation that star cluster formation should scale directly with star formation, this efficiency varies widely across galaxies. To explore this variation we measure the z-band GC specific frequency ($S_{N,z}$) for 43 early-type galaxies (ETGs) from the Hubble Space Telescope (HST)/Advanced Camera for Surveys (ACS) Fornax Cluster Survey. Combined with the homogenous measurements of $S_{N,z}$ in 100 ETGs from the HST/ACS Virgo Cluster Survey from Peng et al. (2008), we investigate the dependence of $S_{N,z}$ on mass and environment over a range of galaxy properties. We find that $S_{N,z}$ behaves similarly in the two galaxy clusters, despite the clusters' order-of-magnitude difference in mass density. The $S_{N,z}$ is low in intermediate-mass ETGs ($-20<M_z<-23$), and increases with galaxy luminosity. It is elevated at low masses, on average, but with a large scatter driven by galaxies in dense environments. The densest environments with the strongest tidal forces appear to strip the GC systems of low-mass galaxies. However, in low-mass galaxies that are not in strong tidal fields, denser environments correlate with enhanced GC formation efficiencies. Normalizing by inferred halo masses, the GC mass fraction, $η=(3.36\pm0.2)\times10^{-5}$, is constant for ETGs with stellar masses $\mathcal{M}_\star \lesssim 3\times10^{10}M_\odot$, in agreement with previous studies. The lack of correlation between the fraction of GCs and the nuclear light implies only a weak link between the infall of GCs and the formation of nuclei.

astro-ph.GA↗

arXiv2017

Virgo Redux: The Masses and Stellar Content of Nuclei in Early-Type Galaxies from Multi-Band Photometry and Spectroscopy

We present an analysis of 39 nuclei and their early-type hosts in the Virgo Cluster using ten broadband filters: F300W, F475W, F850LP, F160W, $u^*griz$, and $K_s$. We describe the Virgo Redux program, which provides high-resolution UV and NIR imaging. Combining this data with optical and NIR imaging from the ACS Virgo Cluster Survey and the Next Generation Virgo Cluster Survey, we estimate masses, metallicities and ages using simple stellar population (SSP) models. For 19 nuclei, we compare to SSP parameters derived from Keck and Gemini spectra and find reasonable agreement between the photometric and spectroscopic metallicity: the RMS scatter is 0.3 dex. We reproduce the nucleus-galaxy mass fraction of $0.33^{+0.09}_{-0.07}$ percent for galaxy stellar masses $10^{8.4}-10^{10.3} M_\odot$ with a typical precision of $\sim$35% for the nuclei masses. Based on available model predictions, there is no single preferred formation scenario for nuclei, suggesting that nuclei are formed stochastically through a mix of processes. Nuclei metallicities are statistically identical to those of their hosts, appearing $0.07 \pm 0.3$ dex more metal-rich on average -- although, omitting galaxies with unusual origins, nuclei are $0.20\pm0.28$ dex more metal-rich. Nuclei appear to be $0.56 \pm 0.12$ dex more metal rich than ultra-compact dwarf galaxies (UCDs) at fixed mass. We find no clear age difference between nuclei and their galaxies, with nuclei displaying a broad range of ages. Interestingly, we find that the most massive nuclei may be flatter and more closely aligned with the semi-major axes of their hosts, suggesting that they formed through predominantly dissipative processes.

astro-ph.GA↗

arXiv2016

The Influence of Environment on the Chemical Evolution in Low-mass Galaxies

The mean alpha-to-iron abundance ratio ([$α$/Fe]) of galaxies is sensitive to the chemical evolution processes at early time, and it is an indicator of star formation timescale ($τ_{\rm SF}$). Although the physical reason remains ambiguous, there is a tight relation between [$α$/Fe] and stellar velocity dispersion ($σ$) among massive early-type galaxies (ETGs). However, no work has shown convincing results as to how this relation behaves at low masses. We assemble 15 data sets from the literature and build a large sample that includes 192 nearby low-mass ($18<σ<80$~\kms) ETGs. We find that the [$α$/Fe]-$σ$ relation generally holds for low-mass ETGs, except in extreme environments. Specifically, in normal galaxy cluster environments, the [$α$/Fe]-$σ$ relation and its intrinsic scatter are, within uncertainties, similar for low-mass and high-mass ETGs. However, in the most massive relaxed galaxy cluster in our sample, the zero point of the relation is higher and the intrinsic scatter is significantly larger. By contrast, in galaxy groups the zero point of the relation offsets in the opposite direction, again with substantial intrinsic scatter. The elevated [$α$/Fe] of low-mass ETGs in the densest environments suggests that their star formation was quenched earlier than in high-mass ETGs. For the low-mass ETGs in the lowest density environments, we suggest that their more extended star formation histories suppressed their average [$α$/Fe]. The large scatter in [$α$/Fe] may reflect stochasticity in the chemical evolution of low-mass galaxies.

astro-ph.GA↗

arXiv2016

The ACS Fornax Cluster Survey. XII. Diffuse Star Clusters in Early-type Galaxies

Diffuse star clusters (DSCs) are old and dynamically hot stellar systems that have lower surface brightness and more extended morphology than globular clusters (GCs). Using the images from HST/ACS Fornax Cluster Survey, we find that 12 out of 43 early-type galaxies (ETGs) in the Fornax cluster host significant numbers of DSCs. Together with literature data from the HST/ACS Virgo Cluster Survey, where 18 out of 100 ETGs were found to host DSCs, we systematically study the relationship of DSCs with GCs, and their host galaxy environment. Two DSC hosts are post-merger galaxies, with most of the other hosts either having low mass or showing clear disk components. We find that while the number ratio of DSCs to GCs is nearly constant in massive galaxies, the DSC-to-GC ratio becomes systematically higher in lower mass hosts. This suggests that DSCs may be more efficient at forming (or surviving) in low density environments. DSC hosts are not special either in their position in the cluster, or in the galactic color-magnitude diagram. Why some disk and low-mass galaxies host DSCs while others do not is still a puzzle, however. The mean ages of DSC hosts and non-hosts are similar at similar masses, implying that formation efficiency, rather than survival, is the reason behind different DSC number fractions in early-type galaxies.

astro-ph.GA↗