Searcharxiv⌕ Search

arXiv subjects

Yuan Fang

Publications and source records attributed to Yuan Fang.

At least 19 recordsLinked to original sources
arXiv2026

From Parameters to Answers: How LLMs Retrieve and Use Their Internal Knowledge

How does a language model's dependence on query-routing information and target knowledge change as it answers a question? We study this question through layerwise interventions on the hidden state at the end of the question. Across Qwen, Llama, and Gemma, we compare country-continent questions with noun, adjective, and code answers while keeping several fitted measurements distinct. A pair-conditioned request direction describes which country is queried in natural single-country questions; a global request direction describes first- versus second-country requests in paired questions; separate selection candidates test control among contents already available in the hidden state. A diagnostic reanalysis of frozen Qwen natural-question states shows that the pair-conditioned direction grows stronger before interventions on it begin to alter later fitted knowledge, with this causal window opening while answer-supporting content is still forming. The paired three-model trajectories are not uniform: Gemma shows a partially overlapping mid-layer routing-content profile, whereas Llama has no sustained routing-effect window under the same gates. In the paired protocol, dependence on the global request direction decreases from fixed earlier to later layer sets while dependence on fitted content persists. A matched Qwen comparison shows that the pair-conditioned direction retains a late effect, so this operational handoff concerns the global fitted direction rather than all request information. These results separate early readability, natural strength, causal steering, and later content dependence.

cs.AI↗
arXiv2026

Rapid Variability and Broadband Spectral Modeling in the Flaring Activity of BL Lacertae

We report a multi-wavelength study of two flaring episodes of the blazar BL Lacertae during MJD 60500-60800 (9 July 2024 - 5 May 2025). The source reached a daily-averaged $\gamma$-ray flux of $(1.03 \pm 0.05) \times 10^{-5} \, \mathrm{ph \, cm^{-2} \, s^{-1}}$ ($E > 100$ MeV) on MJD 60588 (5 October 2024). Using orbit-binned data from the Large Area Telescope (LAT) onboard the \textit{Fermi Gamma-ray Space Telescope}, we identify a minimum flux halving timescale of $\tau = 1.33 \pm 0.29$ hr. This constrains the upper limit on the $\gamma$-ray emitting region size to $R \le 2.0 \times 10^{15}$ cm, as well as its distance from the central supermassive black hole to $R_\mathrm{H} \le 5.9 \times 10^{16}$ cm, assuming a Doppler factor of $\delta = 14.8$ derived from the spectral energy distribution (SED) modeling. We find tentative evidence for sub-minute $\gamma$-ray variability with a minimum doubling time of $0.7 \pm 0.2$ min ($p$-value = 0.03). This may originate from an extremely compact region with a size of $R \le 1.8 \times 10^{13}$ cm, suggesting that the emission arises from magnetohydrodynamic substructures, such as plasmoids within a magnetic reconnection zone. Spectral analysis reveals a significant ``softer-when-brighter'' trend ($r = 0.96, p = 4.5 \times 10^{-4}$) during the minute-scale flare peaks, indicating a complex interplay between particle acceleration and radiative cooling. The SED is reproduced using a one-zone leptonic model, in which synchrotron self-Compton (SSC) and external Compton (EC) scattering effectively account for the high-energy emissions. The reduced magnetic field strengths and hard electron injection spectral indices observed during the flaring states suggest enhanced particle acceleration efficiency, possibly associated with relativistic magnetic reconnection.

astro-ph.HE↗
arXiv2026

TradingMoE: Routing the Right Experts in Evolving Markets

Large language models (LLMs) have shown strong potential for financial analysis and trading, but direct trading remains challenging because the predictive capabilities required can vary across assets, decision fields, and market conditions. Existing LLM-based trading systems either coordinate human-defined external experts or adopt conventional internal Mixture-of-Experts (MoE) routers that do not directly evaluate how individual experts contribute to trading decisions. Moreover, these routers receive no direct signal indicating when an inactive expert has become more suitable as market conditions change. We find that native router scores poorly reflect how much individual experts improve trading decisions, frequently leaving better alternatives unselected. We further reveal that token-specific expert usefulness exhibits a compact low-dimensional structure. Based on these findings, we propose TradingMoE, a trading-oriented sparse MoE that augments a frozen dense LLM with lightweight residual experts. We introduce a Query-Key router that represents the expertise required by each token under the current market context as a low-dimensional query and matches it with learnable expert keys. We further propose a sparse expert selection update mechanism that samples a few inactive experts during training and estimates whether they should replace the weakest expert in the current Top-k route. This mechanism enables the router to update expert selection as market conditions change while preserving sparse computation. Experiments against 22 baselines on stock and cryptocurrency markets show that TradingMoE improves cumulative return over the best-performing baselines by 30.89% and 30.7%, respectively. Rolling paper-trading experiments further demonstrate that its advantage persists under forward-only deployment.

cs.LG↗
arXiv2026

Quantum Fisher information of magnetic quantum phase transition on Kondo lattice

Strange metals exemplify highly collective quantum many-body systems that call for new means of characterization, and there is considerable potential for quantum information approaches contributing to the cause. We investigate multipartite entanglement across the quantum phase transition of a Kondo lattice model using the quantum Fisher information (QFI). We show that the QFI associated with the spin components transverse to the order parameter characterizes the destruction of heavy quasiparticles in the Kondo-destroyed magnetic-ordered phase. The physical origin of this observation is elucidated through an analysis of the antiferromagnetic Heisenberg model. We propose to test the results in terms of both unpolarized and polarized inelastic neutron scattering measurements in the ordered part of the heavy fermion phase diagram. Our findings illustrate how different operators of a many-body system can be employed to not only witness multipartite entanglement in different sectors and but also elucidate the overall physics across different parts of the phase diagram.

cond-mat.str-el↗
arXiv2026

X-rays breaking out of pre-explosion ejecta mark a supernova's first light

Massive stars die as core-collapse supernovae, whose optical light emerges days after the implosion. Theory predicts that the initial collapse-driven shock, upon breaking through the star and dense circumstellar medium, emits a brief thermal flash of soft X-rays and ultraviolet. Yet these elusive first signals have remained largely undetected, owing to limited wide-field soft X-ray monitoring. Here we report the discovery of a soft X-ray flash, EP260321a, followed days later by a broad-lined supernova from an envelope-stripped progenitor. Its X-ray spectrum, best modeled with blackbody, establishes it as the long-sought archetypal shock breakout. The burst's duration and energetics place the breakout at a radius of 300 solar radii, tracing a dense surrounding shell and revealing abrupt mass ejection within the final month before collapse.

astro-ph.HE↗
arXiv2026

Omni-Supervised Motion Editing: Balancing Change and Invariance through Positive-Negative Learning

Text-based human motion editing aims to modify existing motion sequences according to natural language instructions while maintaining the consistency of the original motion. Existing diffusion-based approaches often rely on heuristic similarity cues or coarse global conditioning, leading to motion distortion and suboptimal semantic alignment. The key challenge lies in balancing change (i.e. precisely editing target regions) and invariance (i.e. preserving unedited parts). To handle such challenge, we propose an Omni-Supervised Positive-Negative Learning framework, named OmniME. Our method integrates three complementary components: (1) retrospective feature supervision that enforces coarse-to-fine consistency across transformer layers,(2) motion preservation mechanism that focuses on subtle variations according to the source-target similarity, and (3) triplet-based semantic alignment that strengthens text-motion correspondence. Together, these components form a unified supervision paradigm that balances change and invariance. Extensive experiments on the MotionFix and STANCE Adjustment datasets demonstrate that OmniME achieves state-of-the-art performance in editing alignment, validating the effectiveness of our unified learning framework. Our source codes and models have been released at: https://github.com/rocket-ycyer/OmniME.git

cs.CV↗
arXiv2026

Clustering as Reasoning: A $k$-Means Interpretation of Chain-of-Thought Graph Learning

Chain-of-Thought (CoT) prompting has shown promise in enhancing the reasoning capabilities of large language models (LLMs) on text-attributed graphs (TAGs). This work reframes CoT-based graph learning through the principle of clustering as reasoning, offering a $k$-means interpretation of how iterative reasoning operates over graph-structured data. We observe that existing graph CoT methods rely on disjoint architectures and fixed graph representations, limiting step-by-step semantic-topological interaction and interpretability. To overcome this limitation, we propose a unified framework named KCoT that integrates CoT reasoning with graph representation learning. Our key theoretical result reveals a formal mathematical correspondence between a Transformer block and the $k$-means algorithm, allowing reasoning to be interpreted as iterative assignment and update steps. Based on this insight, we introduce a Semantic Discriminating Prompt that explicitly formulates these steps as structured CoT reasoning, together with a structure-grounded alignment strategy to fuse topological priors with evolving thought-conditioned representations. Experiments on standard benchmarks demonstrate consistent improvements over state-of-the-art methods, validating clustering as a principled mechanism for CoT-based graph learning.

cs.AI↗
arXiv2026

Event-Aware Prompt Learning for Dynamic Graphs

Real-world graph typically evolve via a series of events, modeling dynamic interactions between objects across various domains. For dynamic graph learning, dynamic graph neural networks (DGNNs) have emerged as popular solutions. Recently, prompt learning methods have been explored on dynamic graphs. However, existing methods generally focus on capturing the relationship between nodes and time, while overlooking the impact of historical events. In this paper, we propose EVP, an event-aware dynamic graph prompt learning framework that can serve as a plug-in to existing methods, enhancing their ability to leverage historical events knowledge. First, we extract a series of historical events for each node and introduce an event adaptation mechanism to align the fine-grained characteristics of these events with downstream tasks. Second, we propose an event aggregation mechanism to effectively integrate historical knowledge into node representations. Finally, we conduct extensive experiments on four public datasets to evaluate and analyze EVP.

cs.LG↗
arXiv2026

HoReN: Normalized Hopfield Retrieval for Large-Scale Sequential Model Editing

Large language models encode vast factual knowledge that can become outdated or incorrect after deployment, yet retraining is prohibitively costly. This motivates lifelong model editing, which updates targeted behavior while preserving the rest of the model. Existing editors, both parameter-modifying and parameter-preserving, degrade severely as edits accumulate and struggle to generalize across paraphrases. We propose HoReN, a codebook-based parameter-preserving editor that wraps a single MLP layer with a discrete key-value memory. HoReN treats each codebook entry as both a knowledge key and a Hopfield stored pattern, retrieves edits by angular similarity on the unit hypersphere, and refines queries through damped Hopfield dynamics so paraphrases converge to the correct memory basin while unrelated inputs remain stable. HoReN achieves strong editing performance with consistent gains across diverse benchmarks spanning standard ZsRE, structured WikiBigEdit, and unstructured UnKE evaluations. Moreover, HoReN scales to 50K sequential edits on ZsRE with stable overall performance above 0.93, while prior editors collapse or degrade severely before reaching 10K. Our code is available at https://github.com/ha11ucin8/HoReN.

cs.LG↗
arXiv2026

GraphReAct: Reasoning and Acting for Multi-step Graph Inference

Reasoning-acting frameworks enhance large language models (LLMs) by interleaving reasoning with actions for dynamic information acquisition. However, extending this paradigm to graph learning remains underexplored. Graph data is inherently structured, with information distributed across nodes and edges and encoded through both topology and latent representations. As a result, effective reasoning over graphs requires not only retrieving informative evidence from the graph, but also progressively refining the accumulated context during multi-step inference. In this work, we propose GraphReAct, a graph reasoning-acting framework that enables step-by-step inference over graph-structured data. Specifically, we design a graph-based action space with two complementary retrieval actions: topological retrieval, which captures local structural dependencies, and semantic retrieval, which accesses non-local but relevant evidence in the representation space. These actions dynamically expand the reasoning context. To further support multi-step reasoning, we introduce another type of action, context refinement, which distills and reorganizes accumulated information into a compact representation. By interleaving reasoning with both retrieval and refinement actions, our framework enables a progressive transition from context expansion to compression. Extensive experiments on six benchmark datasets demonstrate that GraphReAct consistently outperforms state-of-the-art methods, validating the effectiveness of reasoning-acting for graph learning.

cs.AI↗
arXiv2026

Nonlinear thermal and thermoelectric transport from quantum geometry

Quantum geometry may enable the development of quantum phases ranging from superconductivity to correlated topological states. One powerful probe of quantum geometry is the nonlinear Hall response which detects Berry curvature dipole in systems with time-reversal invariance and broken inversion symmetry. With broken time-reversal symmetry, this response is also associated with quantum metric dipole. Here we investigate nonlinear thermal and thermoelectric responses, which provide a wealth of new information about quantum geometry. In particular, we uncover a web of connections between these quantities that parallel the standard Wiedemann-Franz and Mott relations. Implications for the studies of a variety of topological systems, including Weyl-Kondo semimetals and Bernal bilayer graphene, are discussed.

cond-mat.mes-hall↗
arXiv2026

CoEvoT: Co-Evolving Chain-of-Thought Prompting for Graph-LLM Reasoning

Graph learning under distribution shift presents a persistent challenge, where models adapt to new graphs with limited or even no supervision. Recent graph--LLM approaches move toward label-efficient prediction by linearizing graphs into prompts and using large language models (LLMs) as predictors, and can adopt Chain-of-Thought (CoT) prompting to exploit LLM's multi-step reasoning capability. However, existing CoT-based graph--LLM methods generate intermediate thoughts while conditioning on fixed graph tokens, limiting step-wise refinement of structural cues. In this paper, we propose CoEvoT, a simple yet effective co-evolving CoT prompting framework for graph--LLM reasoning. CoEvoT couples text-to-graph token rewriting and graph-to-text reasoning guidance in a closed loop: each intermediate textual thought is used to update the graph token evidence state via a lightweight condition network, and the updated tokens are fed back into the next-step instruction to guide subsequent LLM reasoning. This enables step-wise, state-aware evidence refinement, rather than reasoning over a fixed graph snapshot. Extensive experiments on eight datasets demonstrate that CoEvoT consistently outperforms state-of-the-art baselines.

cs.CL↗
arXiv2026

The Intermediate-Mass Black Hole Reverberation Mapping Project: Stable Optical Continuum Lags of an IMBH in the Dwarf Galaxy NGC 4395 Over Years

NGC 4395 is a nearby dwarf spiral galaxy hosting an active galactic nucleus (AGN) powered by an intermediate-mass black hole (IMBH, $M_{\rm BH} \sim 10^{4}$--$10^{5}\,M_\odot$). Recent optical continuum reverberation mapping studies have suggested potential lag variations between different epochs, offering important clues to the physical mechanisms governing variability in the vicinity of the central black hole. We present continuous intranight multi-band photometric monitoring of NGC 4395 based on five nights of observations, including three nights from the Faulkes Telescope North (two of which are archival) and two new nights from Mephisto. This represents the first systematic investigation of optical continuum lag stability in a galaxy hosting a robustly confirmed IMBH. By applying difference-imaging techniques to both the new observations and the reprocessed archival data, we detect statistically significant optical inter-band lags of $\sim 5$--15 minutes, which increase monotonically with increasing wavelength. No obvious $u$-band lag excess is observed, implying a negligible fractional contribution from diffuse continuum (DC) emission to the optical continuum, in agreement with our spectral decomposition results. The inter-band lags remain stable over multi-year baselines. We suggest that this long-term lag stability may be related to the minor DC contribution, a relatively steady disk-corona structure, and the unusually high X-ray-to-optical luminosity ratio characteristic of low-luminosity AGNs, which likely allows X-ray reprocessing to dominate over other potential variability mechanisms. Future facilities like Gemini/SCORPIO, with its simultaneous optical-to-near-infrared coverage, will be ideally suited to play an important role in advancing this field.

astro-ph.GA↗
arXiv2026

A generalised pre-training strategy for deep learning networks in semantic segmentation of remotely sensed images

In the segmentation of remotely sensed images, deep learning models are typically pre-trained using large image databases like ImageNet before fine-tuned on domain-specific datasets. However, the performance of these fine-tuned models is often hindered by the large domain gaps (i.e., differences in scenes and modalities) between ImageNet's images and remotely sensed images being processed. Therefore, many researchers have undertaken efforts to establish large-scale domain-specific image datasets for pre-training, aiming to enhance model performance. However, establishing such datasets is often challenging, requiring significant effort, and these datasets often exhibit limited generaliza-bility to other application scenarios. To address these issues, this study introduces a novel yet simple pre-training strategy designed to guide a model away from learning domain-specific features in a pre-training dataset during pre-training, thereby improving the generalisation ability of the pre-trained model. To evaluate the strategy's effectiveness, deep learning models are pre-trained on ImageNet and subsequently fine-tuned on four semantic segmentation datasets with diverse scenes and modalities, including iSAID, MFNet, PST900 and Potsdam. Experimental results show that the proposed pre-training strategy led to state-of-the-art accuracies on all four datasets, namely 67.4% mIoU for iSAID, 56.9% mIoU for MFNet, 84.22% mIoU for PST900, 91.88% mF1 for Potsdam. This research lays the groundwork for developing a unified foundation model applicable to both computer vision and remote sensing applications.

cs.CV↗
arXiv2026

MAEPose: Self-Supervised Spatiotemporal Learning for Human Pose Estimation on mmWave Video

Millimetre-wave (mmWave) radar offers a more privacy-preserving alternative to RGB-based human pose estimation. However, existing methods typically rely on pre-extracted intermediate representations such as sparse point clouds or spectrogram images, where the rich spatiotemporal information naturally present in radar video streams is discarded for model learning, while such signal processing adds system complexity. In addition, existing solutions are mainly conducted in an end-to-end supervised manner without leveraging unlabelled raw video streams to learn generalized representations. In this study, we present MAEPose, a masked autoencoding-based human pose estimation approach that operates directly on mmWave spectrogram videos. MAEPose learns spatiotemporal motion-aware generalized representations from unlabelled radar video, and leverages its heatmap decoder for multi-frame pose estimation predictions. We evaluate it across three datasets based on leave-one-person-out cross-validation with rigorous statistical testing. MAEPose consistently outperforms state-of-the-art baselines by up to 22.1% in MPJPE p<0.05, and maintains robust accuracy under zero-shot bystander interference with only a 6.5% error increase. Ablation studies confirm that both the pre-training and the heatmap decoder contribute substantially, while modality analysis indicates that leveraging Range-Doppler video as input achieves better pose estimation performance than Range-Azimuth or their fusion, with lower computational cost.

cs.CV↗
arXiv2026

Pathway to lowest-energy structures and stress relaxation for the surface triple junction verified by machine learning

The behavior of surface triple junctions (STJ) at emergent grain boundaries on free surfaces is critical to the microstructure evolution, and therefore to the stability of the next generation interconnect. Yet,despite this significant importance, its lowest-energy structure and local stress have remained persistently unknown. Here, we fill this critical gap through high-resolution experimental mapping of the local surface deformation at STJ, the analysis of the local structure and stress relaxation, and ergodic searching metastable structures. We establish the zipped Y-shaped notch as the universal lowest-energy structures. This energetic preference was well explained by the distinctive local stress mechanism and was excellently verified with machine learning methods for a wide range of boundaries. By revealing the elusive thermodynamics of STJs, our findings advance the research field by redefining the energetic framework for capillary driven structure evolution and providing foundation for understanding kinetically diffusive deformation and for engineering thin-film interconnects and related materials.

cond-mat.mtrl-sci↗
arXiv2026

Reverse Constitutional AI: A Framework for Controllable Toxic Data Generation via Probability-Clamped RLAIF

Ensuring the safety of large language models (LLMs) requires robust red teaming, yet the systematic synthesis of high-quality toxic data remains under-explored. We propose Reverse Constitutional AI (R-CAI), a framework for automated and controllable adversarial data generation that moves beyond isolated jailbreak prompts. By inverting a harmless constitution into a constitution of toxicity and iteratively refining model outputs through a critique--revision pipeline, R-CAI enables scalable synthesis of multi-dimensional adversarial data without human annotation. Optimizing solely for toxicity-related rewards, however, can lead to reward hacking and degraded semantic coherence. To address this challenge, we introduce probability clamping within reinforcement learning from AI feedback, which stabilizes adversarial optimization while preserving adversarial intent. Experiments demonstrate that R-CAI generates diverse, high-quality toxic data and that probability clamping substantially improves semantic coherence (15%) without sacrificing adversarial strength. Overall, R-CAI provides a fully automated framework for red teaming data generation and systematic safety evaluation of aligned language models.

cs.CL↗
arXiv2026

Effect of Rashba spin-orbit coupling on Faraday rotation in an extended Haldane model

Utilization of Faraday rotation (FR) properties of topological materials offers a promising route toward novel magneto-optical devices. We systematically investigated the effect of Rashba spin-orbit coupling (SOC) on FR spectra in an extended Haldane model, which incorporates Rashba SOC and exchange splitting into the original spinless Haldane framework. Using the Kubo formalism, we calculated the FR spectra across the model's rich topological phase diagram. We found that in the Chern number C=2 region, in the absence of exchange splitting, the FR angle can exceed 4$^\circ$ and its peak position is tunable by the Rashba SOC. In contrast, with the inclusion of exchange splitting, a nearly flat FR profile emerges over a broad frequency range, and the FR peak values increase monotonically with the Rashba SOC strength. The Rashba SOC opens additional transition channels, whose net contribution constructively enhances the FR peak. Furthermore, we derived a low-energy effective Hamiltonian expanded up to quadratic terms, the results of which are in good agreement with tight-binding model calculations, thereby validating our numerical results. Our findings suggest that magneto-optical device characteristics can be designed and optimized through Rashba SOC engineering.

cond-mat.mes-hall↗