arXiv subjects

Jiahao Liang

Publications and source records attributed to Jiahao Liang.

At least 19 recordsLinked to original sources

arXiv2026

OneModel: A Unified Foundation for Platform-Scale Multi-Scenario Ranking

Platform-scale recommender systems often span multiple business streams such as organic recommendation, advertising, and merchant services, where user behaviors form a continuous cross-stream trajectory. Maintaining separate ranking systems fragments user representations and increases engineering cost. We propose \textbf{OneModel}, a unified framework for multi-stream final ranking. OneModel maps heterogeneous behaviors into shared event sequences, learns long-context user representations with an action-oriented backbone, and introduces \emph{Scenario-aware Information Modulation} to balance cross-stream transfer and stream-specific specialization. For production deployment, OneModel further adopts stratified user representation, multi-objective training, and optimized online serving with feature decomposition, user feature prefetching, shared user-tower computation, and graph-level inference optimization. We deploy OneModel in production at \emph{Xiaohongshu}, where it delivers consistent offline gains over strong baselines and scales favorably with context length and model capacity. Online A/B tests improve Time Spent by \textbf{+0.33\%} and Engagement by \textbf{+1.25\%} in Explore Feed, lift advertising value by \textbf{+3.43\%} and CTR by \textbf{+8.18\%} in Feed Advertising, and raise DGMV by \textbf{+1.1867\%} and GPM by \textbf{+2.1585\%} in Merchant Recommendation, validating unified multi-stream ranking as an effective production foundation.

cs.IR↗

arXiv2026

Hierarchical Latent Reasoning for LLM-based Recommendation

Large Language Models (LLMs) have shown strong potential for recommendation by leveraging their semantic understanding and contextual modeling capabilities. Recent studies further introduce reasoning mechanisms to improve user preference modeling. However, explicit natural-language reasoning incurs substantial inference overhead, whereas existing latent reasoning methods mainly focus on generating or verifying intermediate states, leaving their layer-wise preference roles and contributions insufficiently characterized. We propose HiLaR, a Hierarchical Latent Reasoning framework with layer-aware reinforcement optimization for LLM-based recommendation. HiLaR constructs temporal-guided hierarchical user preference representations, aligns them with multiple LLM latent reasoning states, and organizes the reasoning process from broad preferences to fine-grained current intents. To further optimize the reasoning trajectory, HiLaR combines final recommendation feedback with layer-aware process rewards derived from the marginal target-likelihood gain of each state. Experiments on four Amazon benchmark datasets show that HiLaR generally outperforms strong sequential, generative, and LLM-based recommendation baselines. Ablation and sensitivity analyses further verify the contribution of hierarchical representation learning, latent alignment, and process-level optimization. Our code is available in https://github.com/hupeiyu21/HiLaR.

cs.IR↗

arXiv2026

Bridging Passive and Active: Enhancing Conversation Starter Recommendation via Active Expression Modeling

Large Language Model (LLM)-driven conversational search is shifting information retrieval from reactive keyword matching to proactive, open-ended dialogues. In this context, Conversation Starters are widely deployed to provide personalized query recommendations that help users initiate dialogues. Conventionally, recommending these starters relies on a closed "exposure-click" loop. Yet, this feedback loop mechanism traps the system in an echo chamber where, compounded by data sparsity, it fails to capture the dynamic nature of conversational search intents shaped by the open world. As a result, the system skews towards popular but generic suggestions. In this work, we uncover an untapped paradigm shift to shatter this harmful feedback loop: harnessing user "free will" through active user expressions. Unlike traditional recommendations, conversational search empowers users to bypass menus entirely through manually typed queries. The open-world intents in active queries hold the key to breaking this loop. However, incorporating them is non-trivial: (1) there exists an inherent distribution shift between active queries and formulated starters. (2) Furthermore, the "non-ID-able" nature of open text renders traditional item-based popularity statistics ineffective for large-scale industrial streaming training. To this end, we propose Passive-Active Bridge (PA-Bridge), a novel framework that employs an adversarial distribution aligner to bridge the distributional gap between passively recommended starters and active expressions. Moreover, we introduce a semantic discretizer to enable the deployment of popularity debiasing algorithms. Online A/B tests on our platform, demonstrate that PA-Bridge significantly boosts the Feature Penetration Rate by 0.54% and User Active Days by 0.04%.

cs.IR↗

arXiv2026

Decoupling momentum and energy relaxation rates in cuprate strange metals via giant THz nonlinearities

Understanding the $T$-linear normal-state resistivity of cuprates remains a central physics challenge. The associated momentum relaxation rate, $Γ_M$, saturates near the conjectured ``Planckian" bound $Γ_M\sim kT/\hbar$, but the mechanism underlying the anomalous scattering remains unresolved. Here we employ nonlinear terahertz spectroscopy to systematically study La$_{2-x}$Sr$_x$CuO$_4$ across a broad temperature and doping range. We measure the normal-state third-order susceptibility, $|χ^{(3)}|\approx 6\times10^{-9}$ m$^2$/V$^2$, among the largest in the THz regime, enabling direct access to the rarely measured electronic energy relaxation rate, $Γ_E$. Strikingly, $Γ_E$ is 10-40 times smaller than $Γ_M$, revealing that the scatterings responsible for momentum loss and $T$-linear resistivity do not remove appreciable energy from the electrons. While $Γ_M (T)$ is consistent with quasi-elastic scattering from bosonic modes above their characteristic energy scale, this is incompatible with the increasing temperature dependence of $Γ_E(T)$. Our results exclude phonons as the source of $T$-linear resistivity and impose strong constraints on possible mechanisms.

cond-mat.supr-con↗

arXiv2026

MuChator: Enabling Active Music Discovery via Conversational Music LLMs in Douyin Music

Douyin Music, a large-scale platform with millions of daily users, adopts an immersive, feed-based discovery paradigm, where users passively explore music through continuous recommendations. While effective for passive music discovery, this paradigm restricts users to recommendation results and provides limited support for explicitly specifying listening intents. Unlike conventional search, where users express well-defined intents through explicit queries such as specific songs or artists, real-world active music discovery is often situational and colloquial, involving vague or underspecified requests. While LLMs enable natural language interaction, their direct use in music discovery remains limited by insufficient music-domain knowledge, lack of music-query collaborative reasoning, and shallow understanding of personalized preferences. To address these challenges, we introduce MuChator, an interactive MusicLLM-based framework that enables users to actively express situational music intents in natural language. MuChator incorporates three key components: (1) Music Knowledge Pre-training, a three-stage scheme that incrementally injects objective music knowledge, subjective music knowledge, and personalized music preferences into LLMs; (2) Context-aware Instruction Tuning, which constructs high-quality user-query-music triplets through an automated synthesis pipeline to align LLMs with active and situational user intents; and (3) Preference Alignment with Hybrid RM, which jointly models intent relevance, personalized preferences, and basic constraints, and is optimized using GRPO-based reinforcement learning. Extensive evaluations on industrial music recommendation datasets demonstrate that MuChator outperforms leading proprietary models, such as Gemini-3-Pro. The model has been deployed on Douyin Music App within ByteDance, with 46.49\% improvement of user active days in online A/B test.

cs.IR↗

arXiv2026

WeatherArchive-Bench: Benchmarking Retrieval-Augmented Reasoning for Historical Weather Archives

Historical archives on weather events are collections of enduring primary source records that offer rich, untapped narratives of how societies have experienced and responded to extreme weather events. These qualitative accounts provide insights into societal vulnerability and resilience that are largely absent from meteorological records, making them valuable for climate scientists to understand societal responses. However, their vast scale, noisy digitized quality, and archaic language make it difficult to transform them into structured knowledge for climate research. To address this challenge, we introduce WeatherArchive-Bench, the first benchmark for evaluating retrieval-augmented generation (RAG) systems on historical weather archives. WeatherArchive-Bench comprises two tasks: WeatherArchive-Retrieval, which measures a system's ability to locate historically relevant passages from over one million archival news segments, and WeatherArchive-Assessment, which evaluates whether Large Language Models (LLMs) can classify societal vulnerability and resilience indicators from extreme weather narratives. Extensive experiments across sparse, dense, and re-ranking retrievers, as well as a diverse set of LLMs, reveal that dense retrievers often fail on historical terminology, while LLMs frequently misinterpret vulnerability and resilience concepts. These findings highlight key limitations in reasoning about complex societal indicators and provide insights for designing more robust climate-focused RAG systems from archival contexts. The constructed dataset and evaluation framework are publicly available at https://anonymous.4open.science/r/WeatherArchive-Bench/.

cs.CL↗

arXiv2026

Reconciling strange metal transport in CeCoIn$_5$ through the difference of optical and cyclotron effective masses

The strange metal behavior in cuprate superconductors - characterized by linear in temperature resistivity and anomalous Hall transport - stands in stark contrast to the expectation of conventional Fermi liquid (FL) theory. Remarkably, the similar transport behavior has also been observed in the heavy fermion metal CeCoIn$_5$, whose d-wave superconducting ground state and strong antiferromagnetic fluctuations draw parallels to the cuprates. Here we have investigated the optical conductivity of the strange metal state of CeCoIn$_5$ over a wide magnetic field range using time-domain THz spectroscopy (TDTS). Using unique high-field THz spectroscopy we have shown that the current relaxation rate scales approximately as T$^2$, giving evidence for a hidden Fermi liquid state over a large field range. This result can be reconciled with linear in T resistivity with the realization that heavy quasiparticles have an optical mass that scales roughly like 1/T. This optical mass contrasts with the mass that characterizes cyclotron motion, which does not suffer the same large temperature dependent renormalization. Although by itself anomalous, this allows one to understand a number of other phenomena in CeCoIn$_5$ that have been taken to be signatures of strange metals, including the coexistence of a conventional T$^2$ dependence of the cotangent of the Hall angle with the linear in T resistivity, which with our observation also reflects FL-like physics.

cond-mat.str-el↗

arXiv2025

Democratic Recommendation with User and Item Representatives Produced by Graph Condensation

The challenges associated with large-scale user-item interaction graphs have attracted increasing attention in graph-based recommendation systems, primarily due to computational inefficiencies and inadequate information propagation. Existing methods provide partial solutions but suffer from notable limitations: model-centric approaches, such as sampling and aggregation, often struggle with generalization, while data-centric techniques, including graph sparsification and coarsening, lead to information loss and ineffective handling of bipartite graph structures. Recent advances in graph condensation offer a promising direction by reducing graph size while preserving essential information, presenting a novel approach to mitigating these challenges. Inspired by the principles of democracy, we propose \textbf{DemoRec}, a framework that leverages graph condensation to generate user and item representatives for recommendation tasks. By constructing a compact interaction graph and clustering nodes with shared characteristics from the original graph, DemoRec significantly reduces graph size and computational complexity. Furthermore, it mitigates the over-reliance on high-order information, a critical challenge in large-scale bipartite graphs. Extensive experiments conducted on four public datasets demonstrate the effectiveness of DemoRec, showcasing substantial improvements in recommendation performance, computational efficiency, and robustness compared to SOTA methods.

cs.IR↗

arXiv2025

Large Language Model Enhanced Graph Invariant Contrastive Learning for Out-of-Distribution Recommendation

Out-of-distribution (OOD) generalization has emerged as a significant challenge in graph recommender systems. Traditional graph neural network algorithms often fail because they learn spurious environmental correlations instead of stable causal relationships, leading to substantial performance degradation under distribution shifts. While recent advancements in Large Language Models (LLMs) offer a promising avenue due to their vast world knowledge and reasoning capabilities, effectively integrating this knowledge with the fine-grained topology of specific graphs to solve the OOD problem remains a significant challenge. To address these issues, we propose {$\textbf{Inv}$ariant $\textbf{G}$raph $\textbf{C}$ontrastive Learning with $\textbf{LLM}$s for Out-of-Distribution Recommendation (InvGCLLM)}, an innovative causal learning framework that synergistically integrates the strengths of data-driven models and knowledge-driven LLMs. Our framework first employs a data-driven invariant learning model to generate causal confidence scores for each user-item interaction. These scores then guide an LLM to perform targeted graph refinement, leveraging its world knowledge to prune spurious connections and augment missing causal links. Finally, the structurally purified graphs provide robust supervision for a causality-guided contrastive learning objective, enabling the model to learn representations that are resilient to spurious correlations. Experiments conducted on four public datasets demonstrate that InvGCLLM achieves significant improvements in out-of-distribution recommendation, consistently outperforming state-of-the-art baselines.

cs.IR↗

arXiv2025

Terahertz range polarization rotation in the candidate time-reversal symmetry breaking superconductor BiNi

Here we report the observation of time-reversal symmetry (TRS) breaking superconductivity in a BiNi bilayer using terahertz (THz) polarimetry. Leveraging a novel high-precision THz polarimetry technique, we detect, in the superconducting state and at zero magnetic field, the smallest polarization rotation of THz light measured to date. By using the MgO substrate itself as an optical resonator, we can reference the Faraday and Kerr rotations to each other. We observe a low-frequency Kerr rotation on the order of several hundred microradians in the superconducting phase, a clear signature consistent with TRS-breaking superconductivity. Our measurements enable direct access to the THz-range Hall conductivity. Through a Kramers-Kronig analysis, we link these low-energy measurements to prior high-frequency magneto-optic Kerr effect (MOKE) data. This connection provides critical insight into the nature of the TRS-breaking state, supporting a multiband superconducting scenario over a disordered single-band interpretation for the origin of the Kerr effect.

cond-mat.supr-con↗

arXiv2025

Stern-Gerlach deflection of cryogenically cold polyatomic molecules in superfluid nanodroplets

Beam deflection is capable of providing valuable information about the magnetic moments of molecules and clusters as well as the relaxation dynamics of their spins. However, observations have been hampered by magnetic couplings to excited vibrational and rotational states of polyatomic systems, which are challenging to control, characterize, and systematize. In this work, we carried out deflection measurements on superfluid helium nanodroplets doped with high-spin FeCl2 and CoCl2 molecules and their complexes. This enabled quantitative determination of the magnetic moments of molecules and clusters at extremely low, and fully defined, temperature of all of their degrees of freedom. The spin magnetic moments become thermalized and oriented along the applied field. Dimers and trimers are found to be antiferromagnetically ordered. The issue of rates and mechanisms of molecular spin relaxation within the cryogenic helium matrix is highlighted.

physics.atm-clus↗

arXiv2025

Amplitude mode in a multi-gap superconductor MgB$_2$ investigated by terahertz two-dimensional coherent spectroscopy

We have investigated the terahertz (THz) nonlinear response of the multigap superconductor MgB$_2$, using THz two-dimensional coherent spectroscopy (THz 2DCS). With broadband THz drive fields, we identified a nonlinear response at twice the lower superconducting gap energy $2Δ_π$ at the lowest temperatures. Using narrow-band THz driving pulses, we observed first (FH) and third harmonic responses. The FH intensity shows a monotonic increase with decreasing temperature when properly normalized by the driving field strength. This is distinct from the single-gap superconductor NbN, where the FH signal exhibited a resonant enhancement at temperatures when twice the gap energy $2Δ$ was resonant with the driving photon energy, which was interpreted to originate from the superconducting amplitude mode. Our results in MgB$_2$ are consistent with a well-defined amplitude mode only at the lowest temperatures and indicate strong damping as temperature increases. This likely indicates the importance of interband coupling in MgB$_2$ and its influence on the nature of the amplitude mode and its damping.

cond-mat.supr-con↗

arXiv2025

Low energy electrodynamics and a hidden Fermi liquid in the heavy-fermion CeCoIn$_5$

We present time-domain THz spectroscopy of thin films of the heavy-fermion superconductor CeCoIn$_5$. Below the $\approx$ 40 K Kondo coherence temperature, a narrow Drude-like peak forms, as the result of the $f$ orbital - conduction electron hybridization and the formation of the heavy-fermion state. The complex optical conductivity is analyzed through a Drude model and extended Drude model analysis. Via the extended Drude model analysis, we measure the frequency-dependent scattering rate ($1/ τ$) and effective mass ($m^*/m_b$). This scattering rate shows a linear dependence on temperature, which matches the dependence of the resistivity as expected. Nevertheless, the width of the low-frequency Drude peak itself that is set by the {\it renormalized} quasiparticle scattering rate ($1 / τ^* = m_b/ m^* τ$) shows a $T^2$ dependence. This is the scattering rate that characterizes the relaxation time of the renormalized quasiparticles. This gives evidence for Fermi liquid state, which in conventional transport experiments is hidden by the strong temperature dependent mass.

cond-mat.str-el↗

arXiv2024

Energy and momentum relaxation through the Curie temperature in an itinerant ferromagnet

In this work, we combine conventional linear response time-domain THz spectroscopy with non-linear THz-pump THz-probe techniques to study metallic strained thin films of $\mathrm{Ca}_2\mathrm{RuO}_4$, which undergo a transition into a ferromagnetic state at 10 K. Such measurements allowing us to independently measure momentum and energy relaxation rates. We find that while the momentum relaxation rate decreases significantly at the ferromagnetic transition, the energy relaxation rate remains unaffected by the emergence of magnetic order. This shows that the dominant changes to scattering across the transition correspond to scatterings that relax momentum without relaxing energy. It is consistent with a scenario where energy is not carried off by coupling to collective magnetic degrees of freedom. Instead, the principal channel for energy relaxation remains the conventional one e.g. coupling to acoustic phonons. This observation validates the approximation used in the conventional understanding of resistive anomalies of ferromagnets across the Curie temperature, which due to critical slowing down, spin fluctuations can be treated as effectively static and scattering off of them elastic. This scenario can likely be extended to resistive anomalies at other phase transitions to charge- and spin-density wave states in kagome metals or pnictide system

cond-mat.str-el↗

arXiv2024

FKA-Owl: Advancing Multimodal Fake News Detection through Knowledge-Augmented LVLMs

The massive generation of multimodal fake news involving both text and images exhibits substantial distribution discrepancies, prompting the need for generalized detectors. However, the insulated nature of training restricts the capability of classical detectors to obtain open-world facts. While Large Vision-Language Models (LVLMs) have encoded rich world knowledge, they are not inherently tailored for combating fake news and struggle to comprehend local forgery details. In this paper, we propose FKA-Owl, a novel framework that leverages forgery-specific knowledge to augment LVLMs, enabling them to reason about manipulations effectively. The augmented forgery-specific knowledge includes semantic correlation between text and images, and artifact trace in image manipulation. To inject these two kinds of knowledge into the LVLM, we design two specialized modules to establish their representations, respectively. The encoded knowledge embeddings are then incorporated into LVLMs. Extensive experiments on the public benchmark demonstrate that FKA-Owl achieves superior cross-domain performance compared to previous methods. Code is publicly available at https://liuxuannan.github.io/FKA_Owl.github.io/.

cs.CL↗

arXiv2024

Mild Solution of Semilinear Rough Stochastic Evolution Equations

In this paper, we investigate a semilinear stochastic parabolic equation with a linear rough term $du_{t}=\left[L_{t}u_{t}+f\left(t, u_{t}\right)\right]dt+\left(G_{t}u_{t}+g_{t}\right)d\mathbf{X}_{t}+h\left(t, u_{t}\right)dW_{t}$, where $\left(L_{t}\right)_{t \in \left[0, T\right]}$ is a family of unbounded operators acting on a monotone family of interpolation Hilbert spaces, $\mathbf{X}$ is a two-step $α$-Hölder rough path with $α\in \left(1/3, 1/2\right]$ and $W$ is a Brownian motion. Existence and uniqueness of the mild solution are given through the stochastic controlled rough path approach and fixed-point argument. As a technical tool to define rough stochastic convolutions, we also develop a general mild stochastic sewing lemma, which is applicable for processes according to a monotone family.

math.PR↗

arXiv2024

Multidimensional Backward Stochastic Differential Equations with Rough Drifts

In this paper, we study a multidimensional backward stochastic differential equation (BSDE) with an additional rough drift (rough BSDE), and give the existence and uniqueness of the adapted solution, either when the terminal value and the geometric rough path are small, or when each component of the rough drift only depends on the corresponding component of the first unknown variable (but dropped is the one-dimensional assumption of Diehl and Friz [Ann. Probab. 40 (2012), 1715-1758]). We also introduce a new notion of the $p$-rough stochastic integral for $p \in \left[2, 3\right)$, and then succeed in giving -- through a fixed-point argument -- a general existence and uniqueness result on a multidimensional rough BSDE with a general square-integrable terminal value, allowing the rough drift to be random and time-varying but having to be linear; furthermore, we connect it to a system of rough partial differential equations.

math.PR↗

arXiv2023

Mild Solution of Semilinear SPDEs with Young Drifts

In this paper, we study a semilinear SPDE with a linear Young drift $du_{t}=Lu_{t}dt+f\left(t, u_{t}\right)dt+\left(G_{t}u_{t}+g_{t}\right)dη_{t}+h\left(t, u_{t}\right)dW_{t}$, where $L$ is the generator of an analytical semigroup, $η$ is an $α$-Hölder continuous path with $α\in \left(1/2, 1\right)$ and $W$ is a Brownian motion. After establishing through two different approaches the Young convolution integrals for stochastic integrands, we introduce the corresponding definition of mild solutions and continuous mild solutions, and give via a fixed-point argument the existence and uniqueness of the (continuous) mild solution under suitable conditions.

math.PR↗