arXiv subjects

Yiding Wang

Publications and source records attributed to Yiding Wang.

At least 19 recordsLinked to original sources

arXiv2026

Spherical Completeness, Coherence, and GCD Properties of Formal Power Series and Witt Vector Rings

Let $K$ be a complete nonarchimedean valued field with $v(K^\times)=\mathbf R$, and let $V=\mathcal O_K$. We prove that $K$ is spherically complete if and only if $V[[T]]$ is coherent, and that this is also equivalent to $V[[T]]$ being a GCD domain. If $K$ is perfect of characteristic $p$, the same characterization holds for the Witt vector ring $W(V)$. Thus, this settles the previously unresolved full-real-value-group case in the coherence problems for both formal power series and Witt vector rings. In particular, this result also gives affirmative answers to Questions~9 and~10 of Anderson--Kang--Park. The proof combines a coherence criterion for complete rings with a spherically complete valuation quotient and a uniform construction of non-finitely generated intersections of two principal ideals from an empty ball chain.

math.AC↗

arXiv2026

ContinualSkillBench: Can LLM Agents Truly Evolve Their Capabilities?

Modern agent frameworks equip large language models with external skill libraries to solve complex tasks. However, it remains unclear whether these systems can effectively evolve their skills and whether the resulting skills improve task-solving capabilities. To bridge this gap, we introduce ContinualSkillBench, a dynamic evaluation framework for in-context continual skill learning. It covers five representative domains, each containing 100 interconnected subtasks ordered by increasing difficulty and opportunities for cross-task skill reuse. Our experiments show that sequential execution generally improves performance, but the gains vary substantially across models and domains. Moreover, in-context learning performs comparably to explicit skill maintenance on average, suggesting that much of the improvement arises from adaptation to prior context and feedback rather than reusable skill abstraction alone. Explicit skills nevertheless provide selective benefits for tasks requiring reusable procedures or precise outputs. We further find that less capable models tend to accumulate larger, more fragmented collections of task-specific skills. These findings show that current in-context skill evolution mechanisms can support continual adaptation, but still struggle to consistently consolidate experience into robust and transferable skills.

cs.AI↗

arXiv2026

Are Common Substructures Transferable? Riemannian Graph Foundation Model with Neural Vector Bundles

Foundation models have sparked a revolution via a pretraining-adaptation paradigm, with recent efforts extending this success to graphs. Unlike other modalities, graphs contain rich structural patterns, yet their structural transferability remains poorly understood. Prior studies consider common substructures in the discrete realm, and we are motivated by a fundamental question: Are common substructures transferable? The underlying theory is largely underexplored. In this work, we shift toward learning transferable structures through the lens of functional behavior. Theoretically, we connect transferable substructures to intrinsic geometry of the representation space. However, characterizing such intrinsic geometry has rarely been touched. Grounded in Riemannian geometry, we develop a graph intrinsic geometry learning framework called Neural Vector Bundle, which enables parsing intrinsic geometry with local coordinates. Building on this, we design GAUGE, a pretrainable neural architecture that constructs the vector bundle, flattening geometrically compatible local coordinates, and a new Dirichlet loss, which also measures the transfer effort. We empirically validate its superior expressiveness in challenging tasks including zero-shot link prediction and graph isomorphism.

cs.LG↗

arXiv2026

Correlations Between Quantum Battery Capacity and Quantum Resources for Two-qubit System

We investigate the relationship between quantum battery capacity and quantum resources in a two-qubit system consisting of mutually coupled battery and charger subsystems. We find that the battery capacity decreases monotonically with the quantum entanglement, steering, Bell nonlocality and coherence, and peaks when these four quantum resources vanish. Moreover, we reveal the capacity gap between the total system capacity and the sum of the battery and charger spin capacities, which is the residual battery capacity, and establish its positive correlation with entanglement. Furthermore, unlike the first four resources, although the battery capacity decreases monotonically with quantum imaginarity, its disappearance under system detuning does not guarantee a peak capacity, and this effect becomes more pronounced as the detuning increases. In contrast to the first five resources, the quantum state texture shows a positive correlation with battery capacity, but a negative correlation with entanglement, steering, Bell nonlocality, coherence, imaginarity, and residual battery capacity. These monotonic relationships are independent of the choice of system parameters. Our findings reveal the relationship between quantum battery capacity and quantum resources during the dynamic evolution of a quantum battery system, and advances the theory of quantum batteries and the development of quantum energy storage systems.

quant-ph↗

arXiv2026

Knowledge is Not Enough: Injecting RL Skills for Continual Adaptation

Large Language Models (LLMs) face the "knowledge cutoff" challenge, where their frozen parametric memory prevents direct internalization of new information. While Supervised Fine-Tuning (SFT) is commonly used to update model knowledge, it often updates factual content without reliably improving the model's ability to use the newly incorporated information for question answering or decision-making. Reinforcement Learning (RL) is essential for acquiring reasoning skills; however, its high computational cost makes it impractical for efficient online adaptation. We empirically observe that the parameter updates induced by SFT and RL are nearly orthogonal. Based on this observation, we propose Parametric Skill Transfer (PaST), a framework that supports modular skill transfer for efficient and effective knowledge adaptation. By extracting a domain-agnostic Skill Vector from a source domain, we can linearly inject knowledge manipulation skills into a target model after it has undergone lightweight SFT on new data. Experiments on knowledge-incorporation QA (SQuAD, LooGLE) and agentic tool-use benchmarks (ToolBench) demonstrate the effectiveness of our method. On SQuAD, PaST outperforms the state-of-the-art self-editing SFT baseline by up to 9.9 points. PaST further scales to long-context QA on LooGLE with an 8.0-point absolute accuracy gain, and improves zero-shot ToolBench success rates by +10.3 points on average with consistent gains across tool categories, indicating strong scalability and cross-domain transferability of the Skill Vector.

cs.LG↗

arXiv2025

Trade-off relations and enhancement protocol of quantum battery capacities in multipartite systems

First, we investigate the trade-off relations of quantum battery capacities in two-qubit system. We find that the sum of subsystem battery capacity is governed by the total system capacity, with this trade-off relation persisting for a class of Hamiltonians, including Ising, XX, XXZ and XXX models. Then building on this relation, we define residual battery capacity for general quantum states and establish coherent/incoherent components of subsystem battery capacity. Furthermore, we introduce the protocol to guide the selection of appropriate incoherent unitary operations for enhancing subsystem battery capacity in specific scenarios, along with a sufficient condition for achieving subsystem capacity gain through unitary operation. Numerical examples validate the feasibility of the incoherent operation protocol. Additionally, for the three-qubit system, we also established a set of theories and results parallel to those for two-qubit case. Finally, we determine the minimum time required to enhance subsystem battery capacity via a single incoherent operation in our protocol. Our findings contribute to the development of quantum battery theory and quantum energy storage systems.

quant-ph↗

arXiv2025

Evaluating Generalization Capabilities of LLM-Based Agents in Mixed-Motive Scenarios Using Concordia

Large Language Model (LLM) agents have demonstrated impressive capabilities for social interaction and are increasingly being deployed in situations where they might engage with both human and artificial agents. These interactions represent a critical frontier for LLM-based agents, yet existing evaluation methods fail to measure how well these capabilities generalize to novel social situations. In this paper, we introduce a method for evaluating the ability of LLM-based agents to cooperate in zero-shot, mixed-motive environments using Concordia, a natural language multi-agent simulation environment. Our method measures general cooperative intelligence by testing an agent's ability to identify and exploit opportunities for mutual gain across diverse partners and contexts. We present empirical results from the NeurIPS 2024 Concordia Contest, where agents were evaluated on their ability to achieve mutual gains across a suite of diverse scenarios ranging from negotiation to collective action problems. Our findings reveal significant gaps between current agent capabilities and the robust generalization required for reliable cooperation, particularly in scenarios demanding persuasion and norm enforcement.

cs.AI↗

arXiv2025

Dual slow-light enhanced photothermal gas spectroscopy on a silicon chip

Integrated photonic sensors have attracted significant attention recently for their potential for high-density integration. However, they face challenges in sensing gases with high sensitivity due to weak light-gas interaction. Slow light, which dramatically intensifies light-matter interaction through spatial compression of optical energy, provides a promising solution. Herein, we demonstrate a dual slow-light scheme for enhancing the sensitivity of photothermal spectroscopy (PTS) with a suspended photonic crystal waveguide (PhCW) on a CMOS-compatible silicon platform. By tailoring the dispersion of the PhCW to generate structural slow light to enhance pump absorption and probe phase modulation, we achieve a photothermal efficiency of 3.6x10-4 rad cm ppm-1 mW-1 m-1, over 1-3 orders of magnitude higher than the strip waveguides and optical fibers. With a 1-mm-long sensing PhCW incorporated in a stabilized on-chip Mach-Zehnder interferometer with a footprint of 0.6 mm2, we demonstrate acetylene detection with a sensitivity of 1.4x10-6 in terms of noise-equivalent absorption and length product (NEAL), the best among the reported photonic waveguide gas sensors to our knowledge. The dual slow-light enhanced PTS paves the way for integrated photonic gas sensors with high sensitivity, miniaturization, and cost-effective mass production.

physics.optics↗

arXiv2025

Law in Silico: Simulating Legal Society with LLM-Based Agents

Since real-world legal experiments are often costly or infeasible, simulating legal societies with Artificial Intelligence (AI) systems provides an effective alternative for verifying and developing legal theory, as well as supporting legal administration. Large Language Models (LLMs), with their world knowledge and role-playing capabilities, are strong candidates to serve as the foundation for legal society simulation. However, the application of LLMs to simulate legal systems remains underexplored. In this work, we introduce Law in Silico, an LLM-based agent framework for simulating legal scenarios with individual decision-making and institutional mechanisms of legislation, adjudication, and enforcement. Our experiments, which compare simulated crime rates with real-world data, demonstrate that LLM-based agents can largely reproduce macro-level crime trends and provide insights that align with real-world observations. At the same time, micro-level simulations reveal that a well-functioning, transparent, and adaptive legal system offers better protection of the rights of vulnerable individuals.

cs.AI↗

arXiv2025

Multi-Agent Evolve: LLM Self-Improve through Co-evolution

Reinforcement Learning (RL) has demonstrated significant potential in enhancing the reasoning capabilities of large language models (LLMs). However, the success of RL for LLMs heavily relies on human-curated datasets and verifiable rewards, which limit their scalability and generality. Recent Self-Play RL methods, inspired by the success of the paradigm in games and Go, aim to enhance LLM reasoning capabilities without human-annotated data. However, their methods primarily depend on a grounded environment for feedback (e.g., a Python interpreter or a game engine); extending them to general domains remains challenging. To address these challenges, we propose Multi-Agent Evolve (MAE), a framework that enables LLMs to self-evolve in solving diverse tasks, including mathematics, reasoning, and general knowledge Q&A. The core design of MAE is based on a triplet of interacting agents (Proposer, Solver, Judge) that are instantiated from a single LLM, and applies reinforcement learning to optimize their behaviors. The Proposer generates questions, the Solver attempts solutions, and the Judge evaluates both while co-evolving. Experiments on Qwen2.5-3B-Instruct demonstrate that MAE achieves an average improvement of 4.54% on multiple benchmarks. These results highlight MAE as a scalable, data-efficient method for enhancing the general reasoning abilities of LLMs with minimal reliance on human-curated supervision.

cs.AI↗

arXiv2025

Beyond Outcome Reward: Decoupling Search and Answering Improves LLM Agents

Enabling large language models (LLMs) to utilize search tools offers a promising path to overcoming fundamental limitations such as knowledge cutoffs and hallucinations. Recent work has explored reinforcement learning (RL) for training search-augmented agents that interleave reasoning and retrieval before answering. These approaches usually rely on outcome-based rewards (e.g., exact match), implicitly assuming that optimizing for final answers will also yield effective intermediate search behaviors. Our analysis challenges this assumption: we uncover multiple systematic deficiencies in search that arise under outcome-only training and ultimately degrade final answer quality, including failure to invoke tools, invalid queries, and redundant searches. To address these shortcomings, we introduce DeSA (Decoupling Search-and-Answering), a simple two-stage training framework that explicitly separates search optimization from answer generation. In Stage 1, agents are trained to improve search effectiveness with retrieval recall-based rewards. In Stage 2, outcome rewards are employed to optimize final answer generation. Across seven QA benchmarks, DeSA-trained agents consistently improve search behaviors, delivering substantially higher search recall and answer accuracy than outcome-only baselines. Notably, DeSA outperforms single-stage training approaches that simultaneously optimize recall and outcome rewards, underscoring the necessity of explicitly decoupling the two objectives.

cs.AI↗

arXiv2025

Improving Quantum Battery Capacity in Tripartite Quantum Systems by Local Projective Measurements

The impact of local von Neumann measurements on quantum battery capacity is investigated in tripartite quantum systems. Two measurement-based protocols are proposed and the concept of optimal local projective operators is introduced. Specifically, explicit analytical expressions are derived for the protocols when applied to general three-qubit X-states. Furthermore, the negative effects of white noise and dephasing noise on quantum battery capacity are analyzed, proving that optimal local projective operators can improve the robustness of subsystem and total system capacity against both noise types for the general tripartite X-state. The performance of different schemes in capacity enhancement are numerically validated through detailed examples and it is found that these optimized operators can effectively enhance both subsystem and total system battery capacity. The results indicate that the local von Neumann measurement is a powerful tool to enhance the battery capacity in multipartite quantum systems.

quant-ph↗

arXiv2025

Notes on detection and measurement of quantum coherence

Quantum coherence is one of the most basic characteristics of quantum mechanics. Here we give some methods to detect and measure quantum coherence. Firstly, we propose a coherence criterion without full quantum state tomography based on partial transposition. Moreover, we present a coherent nonlinear detection strategy from witnesses, in which we find that for some coherent states, normal witness detection fails but our nonlinear detection succeeds. In addition, we prove that when the nonlinear detection on the two copies of the coherent state fails, the nonlinear detection on the three copies may be successful. Finally, due to the difficulty in calculating robustness of coherence for general states, we introduce a lower bound for coherent robustness based on the witness operator, and after comparing our lower bound with the currently known lower bound, one show that our lower bound is better. Coherence is believed to play a crucial role in quantum information tasks, making the detection and quantization of coherence particularly significant. Therefore, these results help to open up new avenues for advancement in quantum theory.

quant-ph↗

arXiv2025

Type III Valley Polarization and Anomalous Valley Hall Effect in Two-Dimensional Non-Janus and Janus Altermagnet Fe2WS2Se2

Exploiting the valley degree of freedom introduces a novel paradigm for advancing quantum information technology. Currently, the investigation on spontaneous valley polarization mainly focuses on two major types of systems. One type magnetic systems by breaking the time-reversal symmetry, the other is ferroelectric materials through breaking the inversion symmetry. Might there be additional scenarios? Here, we propose to realize spontaneous valley polarization by breaking the mirror symmetry in the altermagnets, named type III valley polarization. Through symmetry analysis and first-principles calculations, we confirm that this mechanism is feasible in Non-Janus Fe2WS2Se2. Monolayer Non-Janus and Janus Fe2WS2Se2 are stable Neel-type antiferromagnetic state with the direct band gap semiconductor. More interestingly, their magnetic anisotropy energy exhibits the rare biaxial anisotropy and a four-leaf clover shape in the xy plane, while the xz and yz planes show the common uniaxial anisotropy. This originated from the fourth-order single ion interactions. More importantly, the valley splitting is spontaneously generated in the Non-Janus Fe2WS2Se2 due to the Mxy symmetry breaking, without requiring the SOC effect. Both the Non-Janus and Janus Fe2WS2Se2 exhibit diverse valley polarization and anomalous valley Hall effect properties. In addition, the magnitude and direction of valley polarization can be effectively tuned by the biaxial strain and magnetic field. Our findings not only expand the realization system of spontaneous valley polarization, but also provide a theoretical basis for the high-density storage of valley degrees of freedom.

cond-mat.mtrl-sci↗

arXiv2025

HD-PiSSA: High-Rank Distributed Orthogonal Adaptation

Existing parameter-efficient fine-tuning (PEFT) methods for large language models (LLMs), such as LoRA and PiSSA, constrain model updates to low-rank subspaces, limiting their expressiveness and leading to suboptimal performance on complex tasks. To address this, we introduce High-rank Distributed PiSSA (HD-PiSSA), a distributed PEFT approach that initializes orthogonal adapters across different devices and aggregates their delta updates collectively on W for fine-tuning. Unlike Data Parallel LoRA or PiSSA, which maintain identical adapters across all devices, HD-PiSSA assigns different principal components of the pre-trained weights to each GPU, significantly expanding the range of update directions. This results in over 16x higher effective updated ranks than data-parallel LoRA or PiSSA when fine-tuning on 8 GPUs with the same per-device adapter rank. Empirically, we evaluate HD-PiSSA across various challenging downstream tasks, including mathematics, code generation, and multi-task learning. In the multi-task setting, HD-PiSSA achieves average gains of 10.0 absolute points (14.63%) over LoRA and 4.98 points (6.60%) over PiSSA across 12 benchmarks, demonstrating its benefits from the extra optimization flexibility.

cs.LG↗

arXiv2025

Quantifying quantum-state texture

Quantum-state texture is a newly recognized quantum resource that has garnered attention with the advancement of quantum theory. In this work, we introduce several potential quantum-state texture measure schemes and check whether they satisfy the three fundamental conditions required for a valid quantum-state texture measure. Specifically, the measure induced by the l_1-norm serves as a vital tool for quantifying coherence, but we prove that it cannot be used to quantify quantum state texture. Furthermore, we show that while relative entropy and robustness meet three fundamental conditions, they are not optimal for quantifying quantum-state texture. Fortunately, we still find that there are several measures that can be used as the measure standard of quantum-state texture. Among them, the trace distance measure and the geometric measure are two good measurement schemes. In addition, the two measures based on Uhlmann's fidelity are experimentally friendly and can serve as an ideal definition of quantum-state texture measures in non-equilibrium situations. All these researches on quantum-state texture measure theory can enrich the resource theory framework of quantum-state texture.

quant-ph↗

arXiv2025

Distribution Relationship of Quantum Battery Capacity

The distribution relationship of quantum battery capacity is investigated. First, it is proved that for two-qubit X-states, the sum of the subsystem battery capacities does not exceed the total system's battery capacity, and the conditions are provided under which they are equal. Then define the difference between the total system's and subsystems'battery capacities as the residual battery capacity (RBC) and show that this can be divided into coherent and incoherent components. Furthermore, it is observed that this capacity monogamy relation for quantum batteries extends to general n-qubit X states and any n-qubit X state's battery capacity distribution can be optimized to achieve capacity gain through an appropriate global unitary evolution. Specifically, for general three-qubit X states, stronger distributive relations are derived for battery capacity. Quantum batteries are believed to hold significant potential for outperforming classical counterparts in the future. These findings contribute to the development and enhancement of quantum battery theory.

quant-ph↗

arXiv2025

Witness based nonlinear detection of quantum entanglement

We present a nonlinear entanglement detection strategy which detects entanglement that the linear detection strategy fails. We show that when the nonlinear entanglement detection strategy fails to detect the entanglement of an entangled state with two copies, it may succeed with three or more copies. Based on our strategy, a witness combined with a suitable quanutm mechanical observable may detect the entanglement that can not be detected by the witness alone. Moreover, our strategy can also be applied to detect multipartite entanglement by using the witnesses for bipartite systems, as well as to entanglement concentrations.

quant-ph↗