Searcharxiv⌕ Search

arXiv subjects

Jianfeng Zhang

Publications and source records attributed to Jianfeng Zhang.

At least 19 recordsLinked to original sources
arXiv2026

Pressure-induced Lifshitz and quantum phase transitions in electron-doped cuprate superconductor

We report the first observations of a pressure-induced Lifshitz transition coupled with a quantum phase transition in electron-doped cuprate superconductor Pr0.87LaCe0.13CuO4 (PLCCO), by combining high-pressure electrical resistance, Hall coefficient and synchrotron X-ray diffraction (XRD) measurements at low temperatures. Our low-temperature Hall coefficient measurements reveal that the Hall coefficient decreases continuously and reaches zero at about 10 GPa (critical pressure of Pc1). Upon further compression beyond Pc1, Hall coefficient unexpectedly changes its sign from negative to positive, signaling a reconstruction of the Fermi surface from electron-dominated to hole-dominated topology. Concurrently, the superconducting transition temperature (Tc) exhibits a monotonic suppression, vanishing completely at 17.6 GPa (critical pressure of Pc2), where the system enters a non-superconducting metallic state. Our low temperature XRD measurements unequivocally demonstrate the absence of any structural phase transition across Pc1 and Pc2. Therefore, the sign change in RH at Pc1 is associated with a Lifshitz transition, which is never found in the compressed bulk electron- or hole-doped cuprate superconductors. Moreover, the quantum phase transition observed at Pc2 contrasts sharply with known high-pressure behavior of hole-doped cuprates, uncovering a fundamental difference on how pressure tunes the ground states of electron- versus hole-doped systems. These findings provide crucial insights into the different pressure responses on the interplay among Fermi surface topology, electronic correlations, and superconductivity between these two kinds of cuprate superconductors.

cond-mat.supr-con↗
arXiv2026

Akira: Towards Operating System Architecture for Agile Specialization

Operating Systems (OS) specialization tailors abstractions and privileges to specific use cases, enabling configurable tradeoffs between performance and isolation. However, increasing application diversity makes low-cost and non-invasive agile specialization challenging. First, modular OS designs decompose the system into components and reduce coupling, but lack abstractions separating component implementations from OS configuration. Second, diverse workloads require different isolation strengths, yet existing OSs largely fix isolation models at design time, making isolation adaptation changes costly and invasive. To address these, we present Akira, an OS architecture that enables efficiently adapt OS composition and isolation. Akira introduces the Abstraction and Encapsulation Unit (AEU), which separates component interfaces and isolation requirements from implementations, hiding modular OS internals and enabling non-invasive reconfiguration. Based on the abstraction, Akira introduces the Control Logic Unit (CLU) and the Bridge-based Isolation Unit (BIU) to separately manage component communication and isolation, enabling their independent adaptation without system-wide modifications. Results show that Akira enables rapid and low-cost OS reconfiguration with only a 1.10x slowdown over native FreeRTOS and reduces reconfiguration effort from 1.38 KLOC of intrusive changes to 0.28 KLOC of incremental updates. On Unikraft-based Redis and Nginx, Akira configurations expand the performance-isolation space by up to 10.4x and 7.8x. We publicly release Akira's source code, at https://github.com/JiangRS17/Akira.

cs.OS↗
arXiv2026

Tabby: An Open Pretraining Recipe for Time Series Foundation Models

In this report, we release Tabby, a long context probabilistic time series foundation model, together with a complete and open recipe of how it was built. Tabby adopts an encoder-only patch Transformer architecture and concentrates the contributions on the data and the training procedure. The pretraining corpus combines an extended real-world collection, GIFT-Eval-Pretrain+ and BLAST, with synthetic data from KernelSynth and CauKerV2, an online generator that composes temporal dynamics through randomly sampled structural causal models. Training couples a progressive convergence schedule, which yields reusable intermediate checkpoints, with a deep quantile supervision objective for intermediate layers. The resulting 145M parameter backbone supports contexts of up to 8,192 observations and serves forecasting, classification, and anomaly detection, while a prompt-tuning module further improves in-distribution forecasting performance with the pretrained weights frozen. Tabby achieves competitive zero-shot forecasting performance on GIFT-Eval and the out-of-distribution TIME benchmark, while the same pretrained backbone also supports classification on the UCR Archive and zero-shot anomaly detection on TSB-AD-U. We release training pipeline and model as open source at huawei-noah/trustworthyAI.

cs.LG↗
arXiv2026

Instance-Adaptive Prompts as Context for Time-Series Foundation Models

Longer histories can improve time-series foundation models (TSFMs), but require substantially higher inference cost. We therefore ask whether contextual information can be provided more efficiently through a compact set of learned token embeddings. We introduce PaCTS, which generates a small set of instance-adaptive latent prompts in the form of continuous embedding tokens conditioned on the visible context. These prompts serve as compact context surrogates for frozen TSFMs. PaCTS constructs them from instance-specific global statistics and further refines them with segment-level temporal information, capturing both global characteristics and local temporal variations. The prompt module is jointly trained and deployed across heterogeneous time series with the frozen backbone. Extensive experiments demonstrate the effectiveness of prompts as context, consistently improving forecasting across context lengths and model architectures. With a shorter input context, PaCTS can outperform the same frozen backbone using double context while requiring substantially less inference computation. Compared with weight-space adaptation methods, PaCTS achieves stronger improvements and better out-of-distribution generalization.

cs.LG↗
arXiv2026

PhyMo: A Physical-Field Modality for Multimodal AI4Physics

Multimodal learning is emerging as a powerful paradigm for AI for Physics (AI4Physics), where predicting physical systems requires the joint interpretation of heterogeneous observations, measurements, and domain knowledge. However, existing approaches typically represent physical quantities and governing equations as generic numerical or textual tokens, overlooking the physical constraints that determine their spatiotemporal interactions. To address this limitation, we introduce the \textbf{physical-field modality} and propose \textbf{PhyMo}, a physics-grounded multimodal framework that organizes heterogeneous measurements through PDE-associated operators. PhyMo follows a three-stage learning procedure: the physical-field encoder is first pretrained through field reconstruction under PDE residual supervision, its representations are subsequently aligned with visual embeddings in a shared latent space, and the fused multimodal representations are finally processed by corresponding downstream prediction heads. Experiments on five datasets spanning diverse physical environments show that PhyMo achieves state-of-the-art performance, compared to the strongest baseline on each dataset, demonstrating the superiority of PhyMo on multimodal representation learning in AI4Physics.

cs.LG↗
arXiv2026

DCRL: Decoupling and Coupling Reinforcement Learning via Policy-Reward Manifold Alignment

Reinforcement learning (RL) has emerged as a key paradigm for improving the reasoning capabilities of large language models (LLMs). However, existing reward systems, such as rule-based and reward-model-based, often exhibit issues such as unstable optimization and reward hacking. In this work, we revisit the general reasoning of LLMs from a geometric perspective, conceptualizing it as a coupled manifold composed of three interdependent sub-manifolds: logical deduction, evaluation, and representation. Based on this perspective, response generation in RL can be interpreted as a decoupling process from the evaluation manifold, while reward estimation corresponds to a decoupling process from the logical deduction manifold. The limitations of rule-based and reward-model RL systems can be geometrically interpreted as the mismatch of policy-reward manifolds during RL process. To address the aforementioned misalignment, we propose Decoupling and Coupling Reinforcement Learning (DCRL) framework, which incorporates two key components: (1) a syllogistic logic-based prompt evolution mechanism that dynamically refines reward rubrics to enhance the expressiveness of the reward manifold; and (2) a policy-reward re-coupling mechanism that jointly updates the reward and policy models, ensuring consistent evaluation and mitigating manifold mismatch during training. Theoretical analysis and extensive experiments across multiple reasoning domains demonstrate that DCRL consistently outperforms both rule-based and reward-model baselines. Notably, a Qwen3-4B model trained under DCRL surpasses a Qwen3-32B baseline and approaches the performance of a Qwen3-235B model, highlighting superior effectiveness and generalization in RL.

cs.LG↗
arXiv2026

Local well-posedness of general mean field game master equations

This paper presents a generic approach for establishing mean field game master equations, applicable whenever the mean field equilibrium can be characterized by a McKean-Vlasov forward-backward stochastic differential equation system. The core of our approach is a representation formula for the first-order Lions derivative of the decoupling field of this forward-backward SDE system. We then employ a bootstrap argument to recursively compute its higher-order derivatives. To demonstrate the method's versatility, we establish the local well-posedness for master equations in three distinct models: extended mean field games, mean field games with volatility control, and mean field games with a major player.

math.PR↗
arXiv2026

Learning to Solve Stochastic Controls with Unknown Drifts and Running Rewards: Theory, Algorithms and Convergence

We study continuous-time and possibly high-dimensional stochastic control problems where drift coefficients and running reward functions are unknown. Due to these missing model primitives, we take the exploratory, reinforcement learning (RL) framework of Wang, Zariphopoulou, and Zhou(2020) with relaxed controls and entropy regularization. The objective is to develop theoretically grounded, efficient and scalable RL algorithms to learn both the optimal value functions (which also solve the exploratory HJB equation) and optimal exploratory feedback control policies. When the diffusion coefficients do not contain control, we employ probabilistic representations of both the optimal value function and its gradient based on an auxiliary state process depending only on the diffusion part of the original dynamics. With a delicate analysis on some properly defined mappings and their fixed points, this leads to the introduction of our policy iteration algorithms and their convergence. We demonstrate the performance of our algorithms through various numerical examples. Finally, we study a special control-dependent diffusion case where probability representation of the Hessian is called for.

cs.LG↗
arXiv2026

End-to-End Cell Detection via Instance-aware Graph Modeling

Accurate cell detection and classification are crucial for pathological analysis, directly affecting diagnostic accuracy and treatment planning. To capture complex cellular interactions beyond visual appearance within the tumor microenvironment, several approaches have employed graph neural networks to model spatial and relational patterns among cell nuclei, yielding promising results. However, these methods typically adopt a two-stage paradigm of visual extraction followed by relational modeling, which necessitates separate tuning for each stage, thereby increasing pipeline complexity and hindering end-to-end joint optimization. In this paper, we propose an end-to-end framework for cell detection and classification that jointly models patch-level visual representations and instance-level interactions, which incorporates a dynamic graph construction module and an instance-aware graph network. Specifically, the graph construction module dynamically builds the graph structure using learnable queries derived from patch-level features as cell instance representations, with adjacency defined by integrating feature similarity and spatial distances. The instance-aware graph network performs adaptive instance filtering and feature reorganization, aggregating them over the cell graph into a topological latent state for a selective state-space transition driven by visual cues, fusing appearance and relational evidence. When evaluated on multiple datasets with different staining protocols for cell and nucleus detection, our method significantly outperforms existing approaches in both detection and classification performance. The code will be released at https://github.com/RuochenLiu23/IGM.

cs.CV↗
arXiv2026

FlowTSFM: Turning Encoder Depth into Quantile Transport

Encoder-based time series foundation models (TSFMs) typically rely on deep stacks of independently parameterized Transformer layers, where only the final forecast is supervised and intermediate representations have no explicit predictive role. We introduce FlowTSFM, an encoder architecture that interprets depth as a recurrent transport process: a single Transformer block is iteratively applied with shared parameters, while a quantile-flow objective supervises intermediate states along a prescribed trajectory from a prior distribution toward the final forecast. The objective combines pinball forecasting loss with path-level position matching. With only 38.8M parameters, FlowTSFM achieves competitive performance on GIFT-Eval and TIME, remaining within 1.8-4.6% MASE of stronger baselines while using approximately $3\times$ fewer parameters than a 12-layer Chronos-2 model (119.5M). Beyond accuracy, we introduce CosMean, a scale-free diagnostic measuring whether recurrent updates consistently align toward the final prediction. Under a matched intermediate-state probing protocol, FlowTSFM achieves a CosMean score of 0.919 compared with 0.350 for Chronos-2, suggesting that recurrent parameter sharing combined with path supervision is associated with substantially more structured predictive trajectories at a favorable accuracy-efficiency trade-off.

cs.LG↗
arXiv2026

Enhanced and robust superconductivity in La0.8Sr0.2NiO2 membranes compressed up to 210 GPa

The discovery of superconductivity in infinite-layer nickelate thin films has opened a new frontier for exploring unconventional oxide superconductors beyond the copper oxide family. However, the presence of substrate hampers investigations under very high pressure. Recently, this problem was circumvented by a study of freestanding Nd0.85Sr0.15NiO2 membranes, which revealed that superconducting transition temperature kept increasing as the pressure was ramped up to 91 GPa, without showing signs of saturation. Naturally, one wonders what would happen at even higher pressures. Here, we report that superconductivity in a freestanding La0.8Sr0.2NiO2 membrane persists under applied ultrahigh pressures up to 210 GPa. The superconducting transition onset temperature exhibits a continuous dome-like evolution, increasing from 16 K at ambient pressure to a peak of 74.5 K at 146 GPa, and then gradually decreasing to 57.4 K at 210 GPa. Such robustness of superconductivity against ultrahigh pressure has not been observed in high-Tc oxide superconductors before.

cond-mat.supr-con↗
arXiv2026

Stackelberg Games with a Robust Leader

In this paper we study a Stackelberg game with one leader and multiple followers. Given the leader's control, the followers solve a Nash game with possibly multiple equilibria. We consider a robust leader who considers the worst scenario, namely the followers would select the equilibrium worst for the leader. By using the weak formulation, the problem induces a zero sum game problem with open loop controls, which is time inconsistent. We shall characterize the last problem through an HJB equation on the Wasserstein space of probability measures. The principal-agent problem with one principal and multiple agents can be viewed as a special case of our problem, but with certain constraints.

math.OC↗
arXiv2026

A General Model for Continuous Time Principal-Agent Problem Under Hidden Action

In this paper, we study a general continuous-time Principal-Agent (PA) problem, where the agent privately makes effort and consumption decisions over time under a contract with payment schemes both in continuous time and in lump sums. In particular, we allow the continuous payment process to be a controlled diffusion, which is directly related to the pay-to-performance sensitivity (PPS) in the empirical literature. In solving the agent's problem, we propose a new sufficient condition that directly yields a solution to the agent's problem, without requiring a separate verification step for the solution obtained from the first-order approach. We also present an example which can be solved explicitly.

q-fin.MF↗
arXiv2026

Mantis: Lightweight Foundation Model for Time Series Classification

While foundation models have revolutionized various domains, their application to time series classification remains rather under-explored, with existing literature predominantly focused on forecasting. To bridge this gap, we introduce \textbf{Mantis}, a transformer-based foundation model pre-trained exclusively on synthetic data via self-supervised contrastive learning. We demonstrate that effective tokenization is critical to unlocking the full potential of transformers, proposing a novel token generator unit. Furthermore, we introduce an enhanced test-time methodology that bridges the performance gap between Mantis and strong specialized approaches by leveraging intermediate-layer representations, self-ensembling, and cross-model embedding fusion. Extensive experiments demonstrate that Mantis establishes a new state-of-the-art, outperforming existing foundation models across four diverse dataset collections covering various application domains.

cs.LG↗
arXiv2026

Time Series as Language: A Universal Tokenizer for General-Purpose Time Series Foundation Models

While Next-Token Prediction (NTP) has unified LLM pretraining, its adaptation to unbounded, continuous time series (TS) remains open. To bridge the gap, we introduce UniTok, a universal tokenizer that transforms TS into discrete tokens, and UniTok-FM, a foundation model pretrained via NTP on these tokens. UniTok-FM is a general-purpose foundation model that supports zero-shot and prompt-boosted forecasting, as well as few-shot generation and classification via training-free in-context inference--a capability not achieved by prior works. Technically, UniTok is a vector-quantized autoencoder incorporating prefix normalization for scale stabilization, a progressive-resolution causal architecture for encoding and decoding, and a structure-preserving reconstruction loss for training. UniTok-FM adopts an off-the-shelf LLM architecture without TS-specific modifications. Instead of pretraining on isolated TS, it performs NTP on context windows formed by multiple series with similar patterns, aiming to capture their shared dynamics. Experiments on forecasting, generation, and classification show that a single unified UniTok-FM consistently outperforms statistical and supervised baselines, achieves competitive performance with task-specific foundation models, and uniquely enables training-free in-context inference across tasks.

cs.LG↗
arXiv2026

Agentic Fusion of Large Atomic and Language Models to Accelerate Superconductor Discovery

Artificial intelligence has accelerated materials discovery through high-throughput prediction and generation, yet the decision problem remains a formidable bottleneck. While current AI systems readily propose millions of candidates, navigating the decision regarding a viable experimental target requires resolving multi-dimensional judgments across atomic-scale numerical computation and high-level semantic reasoning. Here we present ElementsClaw, an agentic framework for materials discovery that orchestrates a suite of Large Atomic Model (LAM) tools finetuned from our proposed 1-billion-parameter model Elements for numerical computation, while leveraging Large Language Models (LLMs) for semantic reasoning. Applied to superconductors, ElementsClaw rediscovers 66 experimentally verified superconductors that are absent from the standard SuperCon3D database. Scaling to 2.4 million equilibrium crystals, ElementsClaw identifies 68,000 high-confidence candidates in just 28 GPU hours (https://developer.damo-academy.com/material), expanding known superconducting space by orders of magnitude compared to datasets curated over decades. Guided by the agent's reasoning, we experimentally synthesize and verify four novel superconductors: the motif-guided Zr$_3$ScRe$_8$ ($T_c$ = 6.5 K), the de novo generated HfZrRe$_4$ ($T_c$ = 5.9 K), the structurally reinterpreted Zr$_4$VRe$_7$ ($T_c$ = 3.5 K), and the database-latent Hf$_{21}$Re$_{25}$ ($T_c$ = 2.5 K). Together, our results establish a knowledge integrated, autonomously orchestrated, and experimentally grounded paradigm for materials discovery.

cs.LG↗
arXiv2026

Seed3D 2.0: Advancing High-Fidelity Simulation-Ready 3D Content Generation

We present Seed3D 2.0, an advanced 3D content generation system built on Seed3D 1.0, with substantial improvements across generation fidelity, simulation-ready capabilities, and application coverage. For geometry, a coarse-to-fine two-stage pipeline decouples global structure learning from high-frequency detail recovery, while a locality-aware VAE achieves higher spatial compression and more efficient decoding. For texture and material generation, we replace the cascaded pipeline of Seed3D 1.0 with a unified PBR model that directly generates multi-view albedo and metallic-roughness maps, enhanced by Mixture-of-Experts scaling and VLM-based semantic conditioning for improved material precision and visual fidelity. Beyond single-object generation, Seed3D 2.0 introduces a simulation-ready model suite comprising scene layout planning, part-aware decomposition, and training-free articulation generation, enabling coherent scene construction and part-level physical interaction across physics and graphics engines. A large-scale human preference study against five recent commercial models shows that Seed3D 2.0 achieves consistent win rates of 69.0% to 89.9% in textured 3D asset generation. Seed3D 2.0 is available on https://exp.volcengine.com/ark/vision?_vtm_=0.0.c70961.d701978.0&mode=vision&modelId=doubao-seed3d-2-0-260328&tab=Gen3D

cs.GR↗
arXiv2026

On Information Controls

In this paper we study an optimization problem in which the control is information, more precisely, the control is a $σ$-algebra or a filtration. In a dynamic setting, we establish the dynamic programming principle and the law invariance of the value function. The latter requires a condition slightly stronger than the (H)-hypothesis for the admissible filtration, and enables us to define the value function on $\mathcal P_2(\mathcal P_2(\mathbb R^d))$, the space of laws of random probability measures. By using a new Itô's formula for smooth functions on $\mathcal P_2(\mathcal P_2(\mathbb R^d))$, we characterize the value function of the information control problem by an Hamilton-Jacobi-Bellman equation on this space.

math.OC↗