Search arXiv⌕ Search

arXiv subjects

Peng Cheng

Publications and source records attributed to Peng Cheng.

At least 37 records · Page 2Linked to original sources
arXiv2026

Learning-Aided Short Code Design for ISAC based on MIMO-OFDM

This paper proposes a deep learning (DL)-based coded waveform design for integrated sensing and communications (ISAC), enabling flexible trade-offs between communication reliability and ranging accuracy in short-block transmissions. The proposed scheme is built upon a practical multiple-input multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM) architecture, where the communication channel state information and the angles of the static targets are assumed available at the transmitter. A transformer-based transmitter encodes input information bits directly into ISAC transmit waveforms to jointly optimize the bit error rate (BER) performance and the delay modified Cramer-Rao bound (MCRB). A corresponding transformer-based receiver is adopted at the communication side to recover the transmitted information bits. We further examine the learned codewords for communication-oriented and sensing-oriented designs, revealing that a balanced ISAC waveform naturally exhibits an intermediate structure between these two extremes. Numerical results illustrate these codeword structures and demonstrate that the proposed design provides substantial trade-off gains over conventional schemes based on standard channel coding and modulation.

cs.IT↗
arXiv2026

Quantum Elliptic Cohomology From Four-Dimensional Minimal Supersymmetric Gauge Theories

In this paper, we study non-perturbative vortex partition functions of four-dimensional N=1 supersymmetric gauge theories on the space-time geometry $T^2 \times D^2$, and we propose that these partition functions offer a notion of quantum elliptic cohomology. For particular U(1)-gauge theories we calculate these partition functions explicitly by applying equivariant localization methods to Handsaw quiver varieties that realize for this particular class of U(1)-gauge theories the moduli spaces of the non-perturbative vortex sectors. The determined vortex partition functions are annihilated by difference operators, which are interpreted as Ward identities among N=(0,2) BPS surface defects. Compared to lower dimensional gauge theories with four supercharges, anomalies play an essential role for a consistent formulation of the four-dimensional partition functions. In our proposal towards a mathematical formulation of the vortex partition functions in terms of equivariant elliptic cohomology, the gauge theory anomalies relate to geometric properties of the Thom sheaves corresponding to the relevant quasimap moduli spaces. Motivated by the explicit computations we reflect on the existence of a `virtual structure sheaf' on the moduli space of quasimaps for a general mathematical theory of quantum elliptic cohomology.

hep-th↗
arXiv2026

Modular and Cost-effective Scanning Photocurrent Microscopy System for Sub-micron characterization of 2D optoelectronic devices

Scanning photocurrent microscopy (SPCM) is a powerful technique for probing local optoelectronic phenomena in 2D semiconducting devices. However, commercial setups remain costly, complex and often lack flexibility and adaptability. In this work, we present a home-built SPCM platform built around the retrofitting of a conventional metallographic microscope by coupling it with different light sources (single-mode fiber-coupled lasers and multimode fiber-coupled high-power LEDs), a motorized XY stage, a digital camera and an electronic readout module. This system enables simultaneous acquisition of photocurrent and reflection intensity maps, requiring minimal modifications of the microscope. We reached sub-micron spatial resolution and high imaging fidelity by correlating photocurrent maps with reflection maps, optical micrographs and AFM topography data on different devices fabricated with different materials (InSe, MoS2, WSe2, Gr), on different substrates (Si/SiO2, compact disk). This work provides a reliable, accessible and reproducible high-performance SPCM platform that can be easily implemented in most laboratories for microscale optoelectronic characterization of 2D devices.

cond-mat.mtrl-sci↗
arXiv2026

MSign: An Optimizer Preventing Training Instability in Large Language Models via Stable Rank Restoration

Training instability remains a critical challenge in large language model (LLM) pretraining, often manifesting as sudden gradient explosions that waste significant computational resources. We study training failures in a 5M-parameter NanoGPT model scaled via $μ$P, identifying two key phenomena preceding collapse: (1) rapid decline in weight matrix stable rank (ratio of squared Frobenius norm to squared spectral norm), and (2) increasing alignment between adjacent layer Jacobians. We prove theoretically that these two conditions jointly cause exponential gradient norm growth with network depth. To break this instability mechanism, we propose MSign, a new optimizer that periodically applies matrix sign operations to restore stable rank. Experiments on models from 5M to 3B parameters demonstrate that MSign effectively prevents training failures with a computational overhead of less than 7.0%.

cs.LG↗
arXiv2026

Distribution amplitudes of vector and axial-vector mesons in a nonperturbatively improved symmetry-preserving framework

Using continuum Schwinger-function methods with a nonperturbatively improved, symmetry-preserving kernel, we deliver predictions for the leading-twist light-front distribution amplitudes (DAs) of the $ρ$, $K^\ast$, $a_1(1260)$, $b_1(1235)$, and the unmixed strange partners of the $1^{++}$ and $1^{+-}$ axial-vector (AV) channels, reconstructed from Mellin moments of the associated Bethe--Salpeter wave functions. For vector mesons, polarisation barely affects longitudinal momentum sharing: the longitudinal and transverse DAs are nearly degenerate, and both narrower than the asymptotic distribution in the second-moment sense. The AV sector is different in kind. Charge conjugation compels one projection -- interchanged between the $1^{++}$ and $1^{+-}$ channels -- to vanish at $x=1/2$ and change sign; breaking $SU_F(3)$ symmetry removes this protection, whereupon the zeroth moments become nonzero and the nodes shift from the midpoint. Under a common weighted normalisation, the $1^{+-}$ zeroth moment is $1.63$ times that of the $1^{++}$ channel, and the profile distortion follows the same pattern. A coupling forbidden by charge conjugation in the symmetric limit, $f_{b_1}=0$, becomes $f_{K_1^{+-}}=0.019\,$GeV in the strange channel: an independent measure of the same symmetry breaking, obtained from a current matrix element rather than from the DA reconstruction. What distinguishes the two sectors is thus a symmetry-enforced zero, not the size of the flavour asymmetry.

hep-ph↗
arXiv2026

Scattering Radiation Emitted Near a Black Hole Horizon

We study a family of outgoing electromagnetic radiation states prepared outside but arbitrarily close to the horizon of a Schwarzschild black hole, with frequencies that remain uniformly bounded in the local source frame. As the preparation point approaches the horizon, gravitational redshift drives this radiation toward zero Killing frequency and into a controlled soft regime at any fixed exterior scattering radius. The radiation considered here may be viewed as part of the Hawking radiation spectrum. After including transmission through the Schwarzschild potential, we obtain the leading soft response in a localized region of weak curvature and formulate an inclusive scattering probability that is infrared safe. Through the infrared triangle, this response is related to electromagnetic memory at future null infinity. The conditioned memory record probes angular projections of the zero-frequency component of the transmitted radiation, together with their correlations with the accompanying hard scattering outcome. It can therefore provide an additional asymptotic constraint and a partial fingerprint of the radiation source.

hep-th↗
arXiv2026

SINDI: An Efficient Index for Sparse Vector Approximate Maximum Inner Product Search

Sparse vector Maximum Inner Product Search (MIPS) is crucial in multi-path retrieval for Retrieval-Augmented Generation (RAG). Recent inverted index-based and graph-based algorithms have achieved high search accuracy with practical efficiency. However, their performance in production environments is often limited by redundant distance computations and frequent random memory accesses. Furthermore, the compressed storage format of sparse vectors hinders the use of SIMD acceleration. In this paper, we propose the sparse inverted non-redundant distance index (SINDI), which incorporates three key optimizations: (i) Efficient Inner Product Computation: SINDI leverages SIMD acceleration and eliminates redundant identifier lookups, enabling batched inner product computation; (ii) Memory-Friendly Design: SINDI replaces random memory accesses to original vectors with sequential accesses to inverted lists, substantially reducing memory-bound latency. (iii) Vector Pruning: SINDI retains only the high-magnitude non-zero entries of vectors, improving query throughput while maintaining accuracy. We evaluate SINDI on multiple real-world datasets. Experimental results show that SINDI achieves state-of-the-art performance across datasets of varying scales, languages, and models. On the MsMarco dataset, when Recall@50 exceeds 99%, SINDI delivers single-thread query-per-second (QPS) improvements ranging from 4.2 to 26.4 times compared with SEISMIC and PyANNs. Notably, SINDI has been integrated into Ant Group's open-source vector search library, VSAG.

cs.DB↗
arXiv2026

Personalized w-Event Privacy for Infinite Stream Estimation

In applications such as event monitoring, log analysis, and video querying, $w$-event privacy protects individual data within a sliding time window while supporting accurate stream statistics. Existing studies on infinite data streams mainly assume homogeneous privacy requirements for all users, which cannot capture user-specific privacy preferences. This paper studies personalized $w$-event privacy for private data stream estimation. We first design the Personalized Window Size Mechanism (PWSM), which supports personalized privacy requirements at each time slot. Based on PWSM, we propose Personalized Budget Distribution (PBD) and Personalized Budget Absorption (PBA) to estimate streaming statistics under $\boldsymbol{w}$-Event $\boldsymbol{\mathcal{E}}$ Personalized Differential Privacy (($\boldsymbol{w}$, $\boldsymbol{\mathcal{E}}$)-EPDP). PBD guarantees that the budget reserved for the next time step is no smaller than the budget consumed in the previous release, while PBA improves the current budget by absorbing unused budgets from the previous $k$ time slots and borrowing from the next $k$ time slots. We further develop Dynamic Personalized Budget Distribution (DPBD) and Dynamic Personalized Budget Absorption (DPBA), which allow users to dynamically adjust privacy requirements while satisfying $(τ, \boldsymbol{w}_B, \boldsymbol{w}_F)$-Event $(\boldsymbol{\mathcal{E}}_B, \boldsymbol{\mathcal{E}}_F)$-Personalized Differential Privacy. We prove that all proposed methods achieve the corresponding personalized differential privacy guarantees and derive their error upper bounds. Experiments show that our methods reduce estimation error by at least $53.6\%$ compared with state-of-the-art algorithms.

cs.DB↗
arXiv2026

MERIT: Efficient In-Place Deletion for Dynamic Graph-Based Approximate Nearest Neighbor Indexes

Graph-based indexes have become the dominant approach to approximate nearest neighbor search (ANNS) over high-dimensional data and play a crucial role in real-world applications such as retrieval-augmented generation, recommendation systems, and vector databases. Despite extensive progress in static graph construction and search, efficient in-place deletion remains challenging because obsolete vectors must be removed without allowing stale incoming edges to consume search capacity or expensive graph-wide maintenance to interrupt online services, e.g., retrieval-augmented generation (RAG) and recommendation platforms. To address this problem, we propose MERIT (MST-based Efficient Repair with In-place updaTes), an in-place update framework with three core techniques: (1) bounded search-based recovery that combines a deleted vertex's outgoing neighbors with its readily searchable in-neighbors, (2) $k_r$-Minimum Spanning Tree (MST) local repair that promotes local connectivity while retaining multiple routing choices for graph search, and (3) versioned-edge invalidation that immediately filters all stale incoming edges to the deleted vertex and progressively removes them as adjacency lists are rewritten. Its integration with the hierarchical HNSW index and the single-layer Vamana index demonstrates applicability across distinct graph structures. Extensive experiments on multiple real-world datasets show that MERIT processes deletion at nearly the cost of inserting one vector, achieves up to $3.02\times$--$18.87\times$ faster deletion than state-of-the-art (SOTA) methods, and keeps search recall stable or even improves it as deletions accumulate.

cs.DB↗
arXiv2026

Gravitational partition function under volume constraints

The Euclidean action provides a bridge between gravitational thermodynamics and the partition function. In this work, we further investigate the gravitational partition function under a fixed-volume constraint, generalizing the fixed-volume on-shell geometry in the massless case. Moving beyond this massless configuration, we construct solutions with nonvanishing mass functions, which give rise to a new class of volume-constrained Euclidean geometries (VCEGs). These geometries possess both a boundary and a horizon. However, closer inspection indicates that the boundary is not intrinsic, but rather artificially introduced and can be extended, leading to the extended volume-constrained Euclidean geometries (ECVEGs). The ECVEGs contain two horizons, each generically associated with a conical singularity. Their Euclidean action is given by one quarter of the sum of the areas of the two horizons. In general, the conical singularities at the two horizons cannot be simultaneously eliminated, except at a critical mass $m = m^*$, which defines the critical ECVEG. Configurations with unavoidable conical singularities are naturally interpreted as constrained gravitational instantons. An analysis of their contributions to the partition function, together with their topological properties, reveals a close analogy between the ECVEGs and the Euclidean Schwarzschild--de Sitter static patch. This suggests that the volume constraint effectively plays a role analogous to that of a cosmological constant in semiclassical quantum gravity.

hep-th↗
arXiv2026

AtomBridge: Agentic VLA Inference Plugin for Long-Horizon Tasks in Scientific Experiments

Robotic laboratories play a critical role in autonomous scientific discovery by enabling scalable, continuous experimental execution. Recent vision-language-action (VLA) models offer a promising foundation for robotic laboratories. However, scientific experiments typically involve long-horizon tasks composed of multiple atomic tasks. Existing VLA models may fail to perform composed tasks formed by reordering and composing these known atomic actions. This limitation can arise from a skill-chaining gap caused by robot-state mismatch: the terminal robot state of one skill can fall outside the valid initial-state distribution of the next. To address this challenge, we propose AtomBridge, an Agentic VLA Inference Plugin for Long-Horizon Tasks in Scientific Experiments. AtomBridge attaches at inference time to a VLA policy already fine-tuned on atomic tasks, while keeping its weights fixed. At each task boundary, it uses LLM-based transition reasoning and robotic-action code generation to insert transitional actions between consecutive tasks. This plug-and-play design mitigates the skill-chaining gap caused by robot-state mismatch without additional VLA fine-tuning or demonstrations of composed long-horizon sequences. Across scientific manipulation sequences in simulation and a real-world experimental environment, AtomBridge improves execution continuity and per-step atomic-task success. On 8-step composed tasks, AtomBridge improves full-sequence success by 10%~25%.

cs.RO↗
arXiv2026

ODEWorld: A Continuous Predictive Architecture via Physical-Time Flow

In the physical world we inhabit, space and time are fundamentally continuous. However, existing machine learning paradigms for world modeling are largely confined to discrete-time prediction, thereby exhibiting significant inefficiency in capturing the dynamics of physical world. We introduce Physical-Time Flow (PT-Flow), a novel approach that learns a continuous latent velocity field operating in physical time. Crucially, the underlying dynamics of sequential data are parameterized by an ordinary differential equation (ODE) embedded in a well-structured representation space. Under this paradigm, the prediction of future can be recast as temporal integration via an ODE solver in the compressed latent space. Building upon PT-Flow, we construct ODEWorld, a continuous-time latent world model that is both efficient and versatile. By extracting time-variant features and enforcing ODE properties on both the dynamical representation space and the latent velocity field, ODEWorld effectively addresses the long-standing representation collapse issue in latent world model literature. This also enables high-quality image reconstruction even after long-horizon prediction. Moreover, its continuous nature allows for arbitrary temporal resolution and even backward prediction, which is impossible for most discrete-time models. Lastly, ODEWorld can provide rich planning-oriented information to facilitate downstream policy learning. Comprehensive experiments demonstrate that ODEWorld successfully reconciles planning-conducive dynamics abstraction with visual realism, excelling in both video generation and robotic control. Project page: https://dstate.github.io/odeworld_website/.

cs.LG↗
arXiv2026

QV-PIC: Query-Aware Visual Position-Independent Caching for Efficient RAG Serving

Retrieval-Augmented Generation (RAG) repeatedly prefills identical text chunks across queries, incurring redundant computations. Position-Independent Caching (PIC) mitigates it by reusing precomputed Key-Value (KV) across positions, but its efficiency is constrained by the large volume of text tokens. Rendering text chunks as images can compress the text into fewer visual tokens, but the rendered-image PIC suffers more severe quality degradation than the text PIC. This representation-specific gap primarily arises from contextual mismatches across independently compiled caches and the loss of fine-grained textual evidence during visual compression. Existing PIC repair methods mainly address the former through selective recomputation, but they incur online computation and cannot recover lost textual details. We propose QV-PIC, a query-aware dual-resolution PIC reuse framework guided by model-native templates. Offline, QV-PIC compiles visual caches under the model's native chat-template prefix, improving PIC quality without online recomputation. Online, it preserves global context with low resolution and restores fine-grained textual evidence within a high-resolution budget by cumulative query relevance scores, retaining the efficiency benefit of visual compression. Across six tasks, QV-PIC improves average F1 by 21.6 points over vanilla rendered-image PIC, closes the gap to vanilla text PIC, and surpasses optimized text PIC by 2.58 F1 while reducing TTFT by 17.2\%. Relative to full prefill, it cuts TTFT by 83.8%.

cs.CL↗
arXiv2026

PhyCheck: Fine-Grained Evidence-Grounded Dataset for Physical Law Understanding in Video-LLMs

Embodied intelligence and world models require video understanding systems to go beyond recognizing objects and actions and develop an understanding of physical regularities. However, despite their strong performance on general video understanding tasks, current video-language models still struggle to reliably determine whether an observed event conforms to specific physical laws. Existing benchmarks primarily assess the physical quality of generated videos, providing limited support for systematically evaluating and improving the physical-law understanding of Video Large Language Models (VideoLLMs). To address this gap, we introduce PhyCheck, a video question answering dataset organized at two complementary levels of granularity. The coarse-grained subset asks models to determine whether the phenomenon shown in a video conforms to or violates physical laws, while the fine-grained subset further examines whether models can capture physical details responsible for the violation or compliance. We use these subsets as structured supervision to improve physical understanding. In addition, the dataset contains a diagnostic subset with external causal context that reveal hidden factors affecting physical plausibility, assessing whether models can recalibrate their judgments accordingly. Experiments with Fine-tune Qwen2.5-VL show that training with the proposed data substantially improves the understanding of physical-consistency, while evaluations in the diagnostic subset reveal that current models still have difficulty incorporating additional causal conditions into their decisions. These findings highlight the gap between recognizing surface-level inconsistencies and understanding underlying physical mechanisms, and provide a foundation for evaluating and improving physical understanding in Video-LLMs.

cs.CV↗
arXiv2026

BPS Invariants for Generalized Toric Calabi-Yau Threefolds

We apply topological vertex techniques to Calabi-Yau threefolds dual to brane webs where several 5-branes can end on the same 7-brane. In this context, we determine how topological string partition functions transform under Hanany-Witten transitions and flops, which allows us to track curves and the associated invariants under such transitions. The contributions of parallel external branes form a universal sector invisible to the 5d SCFT; once it is removed, invariants can be transported between different geometries engineering the same theory. This yields an efficient technique to compute Gopakumar-Vafa invariants at any degree and genus. We illustrate this with local Hirzebruch and del Pezzo surfaces, including $dP_4$, whose invariants we obtain at high degree for the first time.

hep-th↗
arXiv2026

SeedPolicy: Horizon Scaling via Self-Evolving Diffusion Policy for Robot Manipulation

Imitation Learning (IL) enables robots to acquire manipulation skills from expert demonstrations. Diffusion Policy (DP) models multi-modal expert behaviors but degrades when naively increasing stacked observation horizons, limiting long-horizon manipulation. We propose Self-Evolving Gated Attention (SEGA), a temporal module that maintains a time-evolving latent state via gated attention, enabling efficient recurrent updates that accumulate long-term context into a compact latent representation while filtering irrelevant temporal information. Integrating SEGA into DP yields Self-Evolving Diffusion Policy (SeedPolicy), which resolves the temporal modeling bottleneck and extends the effective temporal horizon with moderate overhead. On the RoboTwin 2.0 benchmark with 50 manipulation tasks, SeedPolicy outperforms DP and other IL baselines. Averaged across both CNN and Transformer backbones, SeedPolicy achieves 36.8% relative improvement in clean settings and 169% relative improvement in randomized challenging settings over the DP. Compared to vision-language-action models such as RDT with 1.2B parameters, SeedPolicy achieves stronger performance in the clean setting with one to two orders of magnitude fewer parameters, demonstrating strong efficiency. These results establish SeedPolicy as a state-of-the-art imitation learning method for long-horizon robotic manipulation. Code is available at: https://github.com/Youqiang-Gui/SeedPolicy.

cs.RO↗
arXiv2026

RoboInspector: Unveiling the Unreliability of Policy Code for LLM-enabled Robotic Manipulation

Large language models (LLMs) demonstrate remarkable capabilities in reasoning and code generation, enabling robotic manipulation to be initiated with just a single instruction. The LLM carries out various tasks by generating policy code required to control the robot. Despite advances in LLMs, achieving reliable policy code generation remains a significant challenge due to the diverse requirements of real-world tasks and the inherent complexity of user instructions. In practice, different users may provide distinct instructions to drive the robot for the same task, which may cause the unreliability of policy code generation. To bridge this gap, we design \textbf{RoboInspector}, a pipeline to unveil and characterize the unreliability of the policy code for LLM-enabled robotic manipulation from two perspectives: the complexity of the manipulation task and the granularity of the instruction. We perform comprehensive experiments with 216 distinct combinations of tasks, instructions, and LLMs in two prominent frameworks. The \textbf{RoboInspector} identifies four main unreliable behaviors that lead to manipulation failure. We provide a detailed characterization of these behaviors and their underlying causes, giving insight for practical development to reduce unreliability. Furthermore, we introduce a refinement approach guided by failure policy code feedback that improves the reliability of policy code generation by up to 35\% in LLM-enabled robotic manipulation, evaluated in both simulation and real-world environments.

cs.RO↗
arXiv2026

From Noisy Traces to Root Causes: Structural Trajectory Analysis and Causal Extraction for Agent Optimization

The optimization of long-horizon agents increasingly relies on reflection-based mechanisms, where a large language model (LLM) acts as an optimizer to diagnose agent failures and improve agent policies. However, real execution traces are difficult to use directly for optimization: large trace collections are often redundant and heterogeneous, making optimization inefficient and prone to overfitting to low-value failures; meanwhile, each individual trajectory also contains many irrelevant steps, while naive context reduction methods such as truncation or sliding windows can discard causally important evidence and produce misleading optimization signals. To resolve this dilemma, we introduce STRACE (Structural TRajectory Analysis and Causal Extraction), a framework that constructs high signal-noise optimization contexts for more precise and effective optimization. At the batch level, STRACE mines failure patterns to filter redundant traces and retain representative failures; within each selected trace, it performs causal localization over a textual dependency graph to remove non-causal steps and identify the true root-cause module for optimization. Empirical results demonstrate that STRACE significantly outperforms standard context-filtering baselines. Notably, on a challenging formal verification task (VeruSAGE-Bench), it successfully optimizes human-expert designed agents, delivering $1.4\times$ success-rate improvement (42.5% to 58.5%). The code is available at https://github.com/moomight/STRACE .

cs.CL↗