AI Research
arXiv 論文トレンド
cs.AI / cs.LG / 宇宙 / エネルギーなど 8 分野の直近投稿
カテゴリ分布
- Quantum Physics(11)
- Computer Vision(10)
- Machine Learning(4)
- Robotics(3)
- Energy Systems(2)
- AI(2)
- cs.HC(1)
- cond-mat.stat-mech(1)
- physics.ao-ph(1)
- cs.CR(1)
- cs.CL(1)
- cond-mat.str-el(1)
- cs.NI(1)
- Materials(1)
論文一覧
- Machine LearningPropose-Filter-Release: Scaling Noise to Typical Sensitivity via Private Filtering
An estimator can be accurate and stable on typical data, yet sensitive to a few possible records. Differentially private mechanisms commonly scale their noise to the worst-case influence of a single r
- cs.HCMicro-Edits: Computational Support for Fine-Scale Edits in Film
In film editing, every frame matters. While macro, content-level edits shape the high-level narrative of a film, small local edits often control timing and direct visual attention. For example, editor
- Quantum PhysicsReconstructing hidden observables via quantum snapshot thermometry
Projective snapshots are a standard readout tool for quantum simulators, but they directly probe only observables that are diagonal in the measurement basis. We show how local equilibration and an ind
- cond-mat.stat-mechRenormalization for Robust Memory without Synchronization
Local error correction can preserve classical information against noise, but protection under synchronized updates need not survive when sites evolve at independent random times. Establishing stabilit
- Machine LearningRethinking Neural Thickets Without Weight Perturbations
The Neural Thickets hypothesis posits that pretrained language models lie in dense neighborhoods of task-specific experts in weight space, which random Gaussian perturbations can uncover. We reproduce
- Computer VisionRead, Allocate, Route: Role-Coupled Tokenization for Pixel-Space Generative Transformers
Pixel-space generative Transformers typically use one token per image patch, so capturing finer detail requires more tokens and makes global attention increasingly expensive. We observe that finer det
- RoboticsSwapBot: Tool-Centric Robot Manipulation via Interchangeable End-Effectors and Tactile Sensing
Robot manipulation in everyday environments inherently involves substantial tool usage. Seemingly simple household tasks, such as cooking pan-fried tofu, require forceful cutting, delicate brushing, a
- Energy SystemsAdaptive Scenario Generation for KL-DRO Around a Learned Density Model
Distributionally robust optimization (DRO) hedges data-driven two-stage stochastic programs against error in the distribution estimated from a finite record, and standard ambiguity sets differ in how
- Computer VisionLuggage re-identification: A study of strategies towards a real-world application
Luggage re-identification (ReID) is a critical task in video surveillance for forensics, security or border management (suspicious or abandoned luggage recognition), as well as luggage management (los
- Quantum PhysicsComparison of entangling strategies for unitary synthesis
Trapped-ion quantum computers are promising as a platform for realizing quantum computational advantage. They promise long coherence times, near all-to-all connectivity, and potential multiqubit gates
- physics.ao-phFrom extreme weather to agricultural consequences with agentic AI
Extreme weather is forecast days ahead, yet its consequences for agriculture remain hard to predict: a forecast says where a storm will strike, not which crops will suffer or how much is at risk, and
- Computer VisionSonoWorld4D: From One Video to a 4D Audio-Visual Scene via Diffusion Synchronization
Recent progress in visual scene generation has extended beyond static 3D scenes toward the more challenging problem of generating dynamic 4D scenes from minimal input. However, sound, despite being in
- Quantum PhysicsEntanglement diagnostics of stochastic-resetting steady states: theory and measurement on superconducting qubits
Stochastic resetting drives quantum systems into nonequilibrium stationary states unreachable under uninterrupted unitary evolution, and their entanglement depends on whether the reset history is reso
- Machine Learningworld2vec: Mechanistic world models through latent dynamics learning
World models can achieve strong control performance without faithfully capturing the dynamics of their environment, limiting their potential for mechanistic interpretability and scientific discovery.
- Computer VisionProbing Spatial State Representations in Visual Foundation Models
An observer sees only part of the surrounding world, yet must keep track of what lies beyond the current view. Spatial perception therefore involves more than recognizing individual images: it calls f
- RoboticsSymAnt: Grounding Long-Horizon Action Anticipation with Adaptable Symbolic Priors
As robots and intelligent systems increasingly interact with humans, they must anticipate long sequences of future actions from only brief observations. While purely neural models typically require ex
- Quantum PhysicsLocality in Dissipative Ground-State Preparation
Can the ground state of a local Hamiltonian be prepared by local dissipation? We address this question for free-fermion Hamiltonians, using a quadratic Lindbladian with linear jump operators for the d
- Quantum PhysicsLeft-Handed Transmission Lines for the Dynamical Casimir Effect
Left-handed (LH) superconducting transmission lines, built with series capacitors and shunt inductors instead of the series inductors and shunt capacitors of a conventional right-handed (RH) line, hav
- Quantum PhysicsOptimal Diamond Distance Estimation for Unitary Channels
The diamond distance between two quantum channels determines the optimal probability of distinguishing them from a single use. We give a quantum algorithm that estimates the diamond distance between t
- Machine LearningAn ODE-based Sampler for Discrete Diffusion Models
The marginals of a continuous diffusion model can be reproduced by a deterministic ordinary differential equation (ODE), known as the probability flow ODE, which provides a foundation for accelerated
- Computer VisionDifficulty Calibration Across Time for Membership Inference in Diffusion Models
Modern diffusion models are trained on massive datasets that often include copyrighted or private images. Membership inference attacks (MIAs) address this by testing whether a given image was in a mod
- Computer VisionL3G-World: Real-Time Infinite World Generation with Latent 3D Gaussian Memory
Interactive world models generate visual content conditioned on user actions, yet maintaining spatial consistency over long-horizon exploration in real time remains challenging. Existing approaches re
- cs.CRTextSeal2: Distortion-Free LLM Watermarking with Copula Mixing and Memoryless Races
We present TextSeal2, a distortion-free watermark for large language models that keeps its signal under speculative decoding: tokens that replace rejected drafts remain watermarked with the same keys,
- Quantum PhysicsExceptional Points as Witnesses of Non-Gaussian Noise
Efficient control of quantum devices requires the knowledge of their environmental noise, yet the commonly measured power spectrum does not fully characterize non-Gaussian fluctuations of realistic en
- cs.CLWhen to Assist and What to Suggest: Joint Gating and Memory-Grounded Hint Generation in One Compact Decoder
Proactive assistants offer short, memory-grounded hints during a conversation, deciding when to assist and what to suggest. Current systems split this between a gate and a larger generator; we evaluat
- Computer VisionFight World Model: Customizable Combat through Visual Interaction Proxies
A playable, programmable game world must maintain persistent state, enforce executable rules, and support extensible gameplay mechanics. Existing video world models can generate plausible combat, yet
- RoboticsA Conveyor Belt Mechanism for Repositionable Magnetic Coupling Between Continuum Robots
This work proposes a novel conveyor belt mechanism enabling continuous, active repositioning of functional elements along the body of a tendon-driven continuum robot. Up to three conveyor belt mechani
- Computer VisionHairDeck: Animation-Ready Stylized Hair Cards from a Single Image
The hair of stylized characters in games and avatars is built from hair cards: artist-placed ribbons whose geometry is the visible hair, animated by spring bones. Existing image-based methods produce
- Energy SystemsAnalog Realization and Measurement of a Forced Aliev-Panfilov Excitable Oscillator
A single resistor moves a discrete analog realization of the forced Aliev-Panfilov excitable oscillator from a chaotic operating point of the model to a regular one. In this model a rate parameter mul
- cond-mat.str-elAlgebraic theory of abelian fractons and beyond
We develop an algebraic theory of fusion, mobility, and statistics in abelian gapped fracton orders and topological phases. The framework generalizes the algebraic description of abelian anyons to sys
- Quantum PhysicsOptimal detuning and gate time for strongly driven single-qubit gates beyond the rotating-wave approximation
Single-qubit quantum gates are implemented as the time-controlled unitary evolution of a two-level system under harmonic driving. The rotating-wave approximation (RWA) provides closed-form gate parame
- Quantum PhysicsProbing fluctuations in photon emission statistics of spin injected quantum dots with time dependent electronic excitation
Electrically spin-injected ensembles of InGaAs/GaAs quantum dots emit circularly polarized single photons, with the emission helicity determined by the spin polarization of the injected electrons. A f
- Quantum PhysicsToric codes all the way down: decoupling translation-invariant topological CSS codes in higher dimensions
Topological Calderbank-Shor-Steane (CSS) codes in more than two spatial dimensions offer features useful for fault-tolerant quantum computation, such as transversal non-Clifford gates and single-shot
- AICollaborative AI agents for autonomous experimental research with solid-state quantum spins
An autonomous research laboratory could conduct an entire study, from formulating questions to preparing manuscripts. Artificial intelligence (AI) offers one route towards this goal. Recent AI systems
- AIPre-Crime for Agents, at the Last Second: Probing When a Recruited Agent Gives In
When AI agents share channels and credentials, one agent can press another to break its operator's rule, a first step towards collusion. Output filters check only what an agent has already written. Ca
- Computer VisionBeyond Ground Truth: A Heuristic Unsupervised Quality Assurance Engine for Diagnosing Calibrational Blindness in Dental Segmentation
In real-world clinical settings, dental X-ray segmentation using deep learning methods suffers from reliability issues. Even when the segmentation quality is bad, deep learning models show high confid
- cs.NIMultistatic ISAC with Deep Learning-Based Clutter Suppression for 5G and Beyond
Integrated sensing and communications (ISAC) systems using 5G and beyond signals of opportunity enable joint wireless sensing and communication using shared infrastructure and spectrum resources. Whil
- Quantum PhysicsClassical communication over quantum channels: does entanglement help in nature?
Can entangling codewords across independent uses of a quantum channel improve the rate of classical communication? An old conjecture states that there is no such entanglement advantage for all channel
- MaterialsOdd Elasticity in Strain-Gradient Continua: Anisotropy Classification and Non-Hermitian Elastodynamics
Odd elasticity extends continuum mechanics to active nonconservative systems containing internal energy sources, such as robotic active metamaterials with nonreciprocal feedback. Modeling these system
- Computer VisionRADIOlogy: A Single-Pass Approach to Training-Free Zero-Shot Anomaly Detection
Current zero-shot anomaly detection methods often rely on complex prompt-learning techniques or computationally expensive sliding-window strategies to achieve fine-grained visual alignment to detect s