arXiv subjects

Chao Lei

Publications and source records attributed to Chao Lei.

At least 19 recordsLinked to original sources

arXiv2026

Mind the Perspective: Let's Reason Recursively for Theory of Mind

Theory of Mind (ToM) reasoning requires inferring agents' beliefs from partial and asymmetric observations, which remains an open challenge for LLMs. Existing prompting-based approaches improve ToM reasoning through observable-event filtering or temporal belief chains, without explicitly modeling nested beliefs. We introduce RecToM, an inference-time framework for ToM reasoning that models nested beliefs via recursive perspective construction. RecToM constructs each character perspective from the preceding character perspective along the character chain specified by the question, reducing higher-order belief questions to actual-world questions within the final constructed perspective. We further provide a KD45 analysis showing that RecToM's perspective construction induces a well-formed belief modality beyond simple event filtering. Experiments on ToM benchmarks, including Hi-ToM, Big-ToM, and FanToM, across multiple LLM backbones show that RecToM consistently outperforms recent advanced approaches, achieving state-of-the-art performance. Notably, RecToM reaches 100\% accuracy on Hi-ToM with GPT-5.4 and Qwen3.5, a benchmark requiring higher-order ToM reasoning.

cs.AI↗

arXiv2026

SVoT: State-aware Visualization-of-Thought for Spatial Reasoning via Reinforcement Learning

Spatial reasoning remains a challenge for Multimodal Large Language Models (MLLMs), as it requires reliable multi-hop inference over both intermediate states and state transitions. Current studies often leave intermediate states unverified and treat state transitions as implicit processes, which limits reliability in multi-hop spatial reasoning. To address this, we propose State-aware Visualization-of-Thought (SVoT), a reinforcement learning framework that generates interleaved, verifiable intermediate states and visualizations. SVoT integrates transition reasoning chains into the generation processes, enabling the model to verify action preconditions and effects through interleaved textual and visual reasoning. We train SVoT via Group Relative Policy Optimization (GRPO), instantiating verification through reward design and evaluating the efficacy of different fine-grained rewards. As existing benchmarks reduce state transitions to single-variable updates, substantially simplifying the problems, we establish five domains by extending classical environments and introducing two novel domains, Pacman and Gather, that require multi-object interactions and numerical reasoning. These domains support systematic evaluation of multi-hop spatial reasoning with quantitative verification of generated intermediate states and transition reasoning. SVoT with transition-aware supervision achieves state-of-the-art performance across the introduced domains, yielding up to a 65% absolute accuracy gain on out-of-distribution test sets.

cs.AI↗

arXiv2026

Charge density wave in a band insulator

Charge density wave (CDW) implies a periodic modulation of the charge density. Typically observed in metallic systems, CDWs arise from Fermi surface instabilities, resulting in the total or partial gapping of the Fermi surface. Here, we present experimental evidence for a CDW state emerging in a band insulator which has no Fermi surface. The bulk and surface of our material platform, Bi4Br4, is gapped over the entire Brillouin zone. Through topographic and spectroscopic imaging at low temperatures, we unveil an unexpected unidirectional charge modulation in Bi4Br4, breaking the lattice translation symmetry. The CDW develops at temperatures below 40 K and adds an energy gap atop the existing insulating gap of Bi4Br4. Furthermore, our transport measurements reveal nonlinear electrical conduction, a phenomenon conventionally associated with the sliding or phason mode of incommensurate CDWs. These highly unusual observations represent a new type of CDW and demand a new theoretical framework for CDWs.

cond-mat.str-el↗

arXiv2026

PathISE: Learning Informative Path Supervision for Knowledge Graph Question Answering

Knowledge Graph Question Answering (KGQA) aims to answer user questions by reasoning over Knowledge Graphs (KGs). Recent KGQA methods mainly follow the retrieval-augmented generation paradigm to ground Large Language Models~(LLMs) with structured knowledge from KGs. However, training effective models to retrieve question-relevant evidence from KGs typically requires high-quality intermediate supervision signals, such as question-relevant paths or subgraphs, which are time- and resource-intensive to obtain. We propose PathISE, a novel framework for learning high-quality intermediate supervision from answer-level labels. PathISE introduces a lightweight transformer-based estimator that estimates the informativeness of relation paths to construct pseudo path-level supervision. This supervision is then distilled into an LLM path generator, whose generated paths are grounded in the KG to provide compact evidence for inductive answer reasoning. ExtensiveISE experiments on three KGQA benchmarks show that PathISE achieves competitive or state-of-the-art KGQA performance, and provides reusable supervision signals that can enhance existing KGQA models, without relying on costly LLM-refined supervision signals. Our source code is available at https://anonymous.4open.science/r/PathISE-2F87.

cs.AI↗

arXiv2026

TENT: A Declarative Slice Spraying Engine for Performant and Resilient Data Movement in Disaggregated LLM Serving

Modern GPU clusters rely on complex, heterogeneous interconnects. As large language model (LLM) serving shifts toward agentic reasoning, KVCache becomes a first-class mobile asset, driving frequent migrations and massive elephant flows that dominate the execution critical path. Operating Mooncake Transfer Engine (TE) on thousands of GPUs exposed a fundamental flaw in existing frameworks: imperative, early-binding path selection. This rigidity results in state-blind striping that ignores congestion and grey failures, leading to bandwidth stranding. This also produces operational fragility where routine faults require manual intervention. We present TENT, a declarative orchestration engine that decouples transfer intent from physical execution. By abstracting interconnects into a unified resource pool, TENT shifts path resolution from initialization to slice-time late binding. Applications simply declare transfer intents, while TENT dynamically "sprays" fine-grained slices across rails based on real-time telemetry and predictive cost modeling. This orchestration eliminates head-of-line (HoL) blocking and enables transparent, sub-50 ms self-healing by rerouting slices around failures or degradations without application-level intervention. TENT serves as the production data plane for LLM inference and reinforcement learning (RL) pipelines at multiple industrial clusters. Our evaluation shows that TE+ outperforms state-of-the-art baselines, including Mooncake TE, NIXL, and UCCL. In LLM inference with SGLang HiCache, TE+ achieves up to 1.36x higher throughput and 26% lower P90 time-to-first-token (TTFT) than Mooncake TE. In RL pipelines, TENT accelerates parameter updates in Moonshot Checkpoint Engine by 20-26%.

cs.DC↗

arXiv2026

MIDI-LLaMA: An Instruction-Following Multimodal LLM for Symbolic Music Understanding

Recent advances in multimodal large language models (MLLM) for audio music have demonstrated strong capabilities in music understanding, yet symbolic music, a fundamental representation of musical structure, remains unexplored. In this work, we introduce MIDI-LLaMA, the first instruction-following MLLM for symbolic music understanding. Our approach aligns the MIDI encoder MusicBERT and Llama-3-8B via a two-stage pipeline comprising feature alignment and instruction tuning. To support training, we design a scalable annotation pipeline that annotates GiantMIDI-Piano with fine-grained metadata, resulting in a MIDI-text dataset. Compared with the baseline trained on converting MIDI into ABC notation under the same instruction-tuning procedure, MIDI-LLaMA substantially outperforms in captioning and semantic alignment in question answering. Human evaluation further confirms the advantages of MIDI-LLaMA in music understanding, emotion recognition, creativity, and overall preference. These findings demonstrate that incorporating symbolic music into large language models enhances their capacity for musical understanding.

cs.MM↗

arXiv2025

PROPA: Toward Process-level Optimization in Visual Reasoning via Reinforcement Learning

Despite significant progress, Vision-Language Models (VLMs) still struggle with complex visual reasoning, where multi-step dependencies cause early errors to cascade through the reasoning chain. Existing post-training paradigms are limited: Supervised Fine-Tuning (SFT) relies on costly step-level annotations, while Reinforcement Learning with Verifiable Rewards (RLVR) methods like GRPO provide only sparse, outcome-level feedback, hindering stable optimization. We introduce PROPA (Process-level Reasoning Optimization with interleaved Policy Alignment), a novel framework that integrates Monte Carlo Tree Search (MCTS) with GRPO to generate dense, process-level rewards and optimize reasoning at each intermediate step without human annotations. To overcome the cold-start problem, PROPA interleaves GRPO updates with SFT, enabling the model to learn from both successful and failed reasoning trajectories. A Process Reward Model (PRM) is further trained to guide inference-time search, aligning the test-time search with the training signal. Across seven benchmarks and four VLM backbones, PROPA consistently outperforms both SFT- and RLVR-based baselines. It achieves up to 17.0% gains on in-domain tasks and 21.0% gains on out-of-domain tasks compared to existing state-of-the-art, establishing a strong reasoning and generalization capability for visual reasoning tasks. The code isavailable at: https://github.com/YanbeiJiang/PROPA.

cs.CV↗

arXiv2025

DUAL: Learning Diverse Kernels for Aggregated Two-sample and Independence Testing

To adapt kernel two-sample and independence testing to complex structured data, aggregation of multiple kernels is frequently employed to boost testing power compared to single-kernel tests. However, we observe a phenomenon that directly maximizing multiple kernel-based statistics may result in highly similar kernels that capture highly overlapping information, limiting the effectiveness of aggregation. To address this, we propose an aggregated statistic that explicitly incorporates kernel diversity based on the covariance between different kernels. Moreover, we identify a fundamental challenge: a trade-off between the diversity among kernels and the test power of individual kernels, i.e., the selected kernels should be both effective and diverse. This motivates a testing framework with selection inference, which leverages information from the training phase to select kernels with strong individual performance from the learned diverse kernel pool. We provide rigorous theoretical statements and proofs to show the consistency on the test power and control of Type-I error, along with asymptotic analysis of the proposed statistics. Lastly, we conducted extensive empirical experiments demonstrating the superior performance of our proposed approach across various benchmarks for both two-sample and independence testing.

cs.LG↗

arXiv2025

Near room temperature magnetoelectric response and tunable magnetic anisotropy in the two-dimensional magnet 1T-CrTe2

Magnets with controllable magnetization and high critical temperature are essential for practical spintronics devices, among which the two-dimensional 1T-CrTe2 stands out because of its high experimental critical temperature up to about 300K down to the single layer limit. By using ab initio density functional theory, we investigate the magnetic properties of monolayer and bilayer 1T-CrTe2 and demonstrate that the magnetic properties, such as the magnetocrystalline anisotropy, critical Curie temperature and magnetizations, can be influenced by strain or electric fields.

cond-mat.mtrl-sci↗

arXiv2025

Beyond Perception: Evaluating Abstract Visual Reasoning through Multi-Stage Task

Current Multimodal Large Language Models (MLLMs) excel in general visual reasoning but remain underexplored in Abstract Visual Reasoning (AVR), which demands higher-order reasoning to identify abstract rules beyond simple perception. Existing AVR benchmarks focus on single-step reasoning, emphasizing the end result but neglecting the multi-stage nature of reasoning process. Past studies found MLLMs struggle with these benchmarks, but it doesn't explain how they fail. To address this gap, we introduce MultiStAR, a Multi-Stage AVR benchmark, based on RAVEN, designed to assess reasoning across varying levels of complexity. Additionally, existing metrics like accuracy only focus on the final outcomes while do not account for the correctness of intermediate steps. Therefore, we propose a novel metric, MSEval, which considers the correctness of intermediate steps in addition to the final outcomes. We conduct comprehensive experiments on MultiStAR using 17 representative close-source and open-source MLLMs. The results reveal that while existing MLLMs perform adequately on basic perception tasks, they continue to face challenges in more complex rule detection stages.

cs.CV↗

arXiv2025

From Reasoning to Generalization: Knowledge-Augmented LLMs for ARC Benchmark

Recent reasoning-oriented LLMs have demonstrated strong performance on challenging tasks such as mathematics and science examinations. However, core cognitive faculties of human intelligence, such as abstract reasoning and generalization, remain underexplored. To address this, we evaluate recent reasoning-oriented LLMs on the Abstraction and Reasoning Corpus (ARC) benchmark, which explicitly demands both faculties. We formulate ARC as a program synthesis task and propose nine candidate solvers. Experimental results show that repeated-sampling planning-aided code generation (RSPC) achieves the highest test accuracy and demonstrates consistent generalization across most LLMs. To further improve performance, we introduce an ARC solver, Knowledge Augmentation for Abstract Reasoning (KAAR), which encodes core knowledge priors within an ontology that classifies priors into three hierarchical levels based on their dependencies. KAAR progressively expands LLM reasoning capacity by gradually augmenting priors at each level, and invokes RSPC to generate candidate solutions after each augmentation stage. This stage-wise reasoning reduces interference from irrelevant priors and improves LLM performance. Empirical results show that KAAR maintains strong generalization and consistently outperforms non-augmented RSPC across all evaluated LLMs, achieving around 5% absolute gains and up to 64.52% relative improvement. Despite these achievements, ARC remains a challenging benchmark for reasoning-oriented LLMs, highlighting future avenues of progress in LLMs.

cs.AI↗

arXiv2025

State-Based Disassembly Planning

It has been shown recently that physics-based simulation significantly enhances the disassembly capabilities of real-world assemblies with diverse 3D shapes and stringent motion constraints. However, the efficiency suffers when tackling intricate disassembly tasks that require numerous simulations and increased simulation time. In this work, we propose a State-Based Disassembly Planning (SBDP) approach, prioritizing physics-based simulation with translational motion over rotational motion to facilitate autonomy, reducing dependency on human input, while storing intermediate motion states to improve search scalability. We introduce two novel evaluation functions derived from new Directional Blocking Graphs (DBGs) enriched with state information to scale up the search. Our experiments show that SBDP with new evaluation functions and DBGs constraints outperforms the state-of-the-art in disassembly planning in terms of success rate and computational efficiency over benchmark datasets consisting of thousands of physically valid industrial assemblies.

cs.RO↗

arXiv2024

Moir\'{e} band theory for M-valley twisted transition metal dichalcogenides

We propose twisted bilayers of certain group IV and IVB trigonal transition metal dichalcogenides (TMDs) MX$_{2}$ (M$=$Zr, Hf, Sn and X$=$S, Se) as moir\'{e} materials. In monolayer form these TMDs have conduction band minima near the three inequivalent Brillouin zone $M$ points and negligible spin-orbit coupling, implying six flavors of low-energy conduction band states. The flavor sectors decouple at the single-particle level and in twisted bilayers are accurately described by emergent moir\'e-periodic Hamiltonians that we derive from small-unit-cell density functional theory calculations. Because the valley-projected Hamiltonians have large valley-dependent mass anisotropies and are time-reversal invariant, spontaneous valley polarization is signaled in transport by anisotropy instead of by the anomalous Hall and magnetic circular dichroism signals commonly observed in graphene and $K$-valley TMD-based moir\'{e} multilayers.

cond-mat.mes-hall↗

arXiv2024

Planning-Driven Programming: A Large Language Model Programming Workflow

The strong performance of large language models (LLMs) raises extensive discussion on their application to code generation. Recent research suggests continuous program refinements through visible tests to improve code generation accuracy in LLMs. However, these methods suffer from LLMs' inefficiency and limited reasoning capacity. In this work, we propose an LLM programming workflow (LPW) designed to improve both initial code generation and subsequent refinements within a structured two-phase workflow. Specifically, the solution generation phase formulates a solution plan, which is then verified through visible tests to specify the intended natural language solution. Subsequently, the code implementation phase drafts an initial code according to the solution plan and its verification. If the generated code fails the visible tests, the plan verification serves as the intended solution to consistently inform the refinement process for correcting bugs. Compared to state-of-the-art methods across various existing LLMs, LPW significantly improves the Pass@1 accuracy by up to 16.4% on well-established text-to-code generation benchmarks. LPW also sets new state-of-the-art Pass@1 accuracy, achieving 98.2% on HumanEval, 84.8% on MBPP, 59.3% on LiveCode, 62.6% on APPS, and 34.7% on CodeContest, using GPT-4o as the backbone. Our code is publicly available at: https://github.com/you68681/lpw

cs.SE↗

arXiv2024

All-optical magnetic imaging with spin defects in van der Waals materials at Angstrom-scale

Magnetic imaging with ultra-high spatial resolution is crucial to exploring the magnetic textures of emerging quantum materials. We propose a novel magnetic imaging protocol that achieves Angstrom-scale resolution by combining spin defects in van der Waals materials and terahertz scattering scanning near-field optical microscopy (THz s-SNOM). Spin defects in the atomic monolayer enable the probe-to-sample distance diving into the Angstrom range where the exchange interactions between the probe and sample spins become predominant. This exchange interaction leads to energy splitting of the probe spin in the order of millielectronvolts, corresponding to THz frequencies. With THz optics and the spin-dependent fluorescence of the probe spin, the interaction energy can be resolved entirely through optical methods. Our proposed all-optical magnetic imaging protocol holds significant promise for investigating magnetic textures in condensed matter physics due to its excellent compatibility and high spatial resolution.

quant-ph↗

arXiv2024

Bulk versus surface: Nonuniversal partitioning of the topological magnetoelectric effect

The electronic ground state of a three-dimensional (3D) band insulator with time-reversal ($Θ$) symmetry or time-reversal times a discrete translation ($ΘT_{1/2}$) symmetry is classified by a $\mathbb{Z}_{2}$-valued topological invariant and characterized by quantized magnetoelectric response. Here we demonstrate by explicit calculation in model $\mathbb{Z}_{2}$ topological insulator thin-films that whereas the magnetoelectric response is localized at the surface in the $Θ$ symmetry (nonmagnetic) case, it is nonuniversally partitioned between surface and interior contributions in the $ΘT_{1/2}$ (antiferromagnetic) case, while remaining quantized. Within our model the magnetic field induced polarization arises entirely from an anomalous $\mathscr{N}=0$ Landau level subspace within which the projected Hamiltonian is a generalized Su-Schrieffer-Heeger model whose topological properties are consistent with those of the starting 3D model. We identify a new connection between the ground-state geometry of that 3D model and surface-interior partitioning in thin films.

cond-mat.mes-hall↗

arXiv2024

Kerr, Faraday, and Magnetoelectric Effects in MnBi$_2$Te$_4$ Thin Films

The topological magneto-electric effect (TME) is a characteristic property of topological insulators. In this article, we use a simplified coupled-Dirac-cone electronic structure model to theoretically evaluate the THz and far infrared Kerr and Faraday responses of thin films of MnBi$_2$Te$_4$ with up to $N=10$ septuple layers with the goal of clarifying the relationship between these convenient magneto-optical observables and the TME. We find that for even $N$ the linear Kerr and Faraday responses to an electric field vanish in the low-frequency limit, even though the magnetoelectric response is large and approximately quantized.

cond-mat.mes-hall↗

arXiv2024

Capacitive scheme to detect the topological magnetoelectric effect

The topological magnetoelectric effect (TME) is a defining property of three-dimensional $\mathbb{Z}_{2}$ topological insulators that was predicted on theoretical grounds more than a decade ago, but has still not been directly measured. In this Letter we propose a strategy for direct measurement of the TME and discuss the precision of the effect in real devices with charge and spin disorder.

cond-mat.mes-hall↗