Searcharxiv⌕ Search

arXiv subjects

Yingying Guo

Publications and source records attributed to Yingying Guo.

16 recordsLinked to original sources
arXiv2026

RoboSynChallenge: Mastering Real-World Dexterity via Generalizing Synthesized Manipulation Skills

Achieving generalizable robotic manipulation remains a central challenge in embodied intelligence. Despite rapid advances in model architectures and learning algorithms, progress is often limited by the scarcity and narrow diversity of real-world data. The RoboSynChallenge competition introduces a unified benchmark to evaluate and advance the generalizability of manipulation policies across a spectrum of tasks, environments, and difficulty levels. To alleviate the shortage of realistic data, the challenge integrates large-scale synthetic data generation with standardized real-world robotic evaluation. Participants are encouraged to leverage synthesized state-action trials to improve general-purpose policy learning, while final assessments are conducted exclusively on unseen real-world manipulation environments. Baseline implementations, including Transformer-, Diffusion-, Vision-Language-Action, and World-Action-Model-based policies, are provided to ensure reproducibility and comparability. By coupling scalable simulation-based training with rigorous real-world validation, RoboSynChallenge aims to foster the development of broadly capable, data-efficient, and adaptable manipulation systems, thereby paving the way toward truly general robotic intelligence.

cs.RO↗
arXiv2026

Experience-Sensitive Game Learning: A Behavioral Study of Humans and Language Agents

Large language model agents are increasingly evaluated through games, but most benchmarks emphasize final outcomes rather than how players learn from repeated interaction. We study experience-sensitive game learning: how gameplay experience changes the decision-making behavior of humans and language agents. We formulate experience-sensitive game learning as a framework for analyzing behavioral change across repeated gameplay, rather than only final score or win rate. We introduce a suite of interactive games with reusable strategic structure, together with cross-game greedy-to-global metrics and game-specific behavioral diagnostics that make experience-driven change observable from action traces. We also collect repeated-game trajectories from human players and evaluate recent self-evolving language agents in the same behavioral metric space. Our results show that human players exhibit interpretable and relatively stable shifts from locally greedy heuristics toward more global strategic decisions. In contrast, current self-evolving agents often show noisy and transient gains, suggesting that existing self-evolution methods remain limited in converting gameplay experience into durable changes in decision-making behavior.

cs.HC↗
arXiv2026

Mind the Unseen Mass: Unmasking LLM Hallucinations via Soft-Hybrid Alphabet Estimation

This paper studies uncertainty quantification for large language models (LLMs) under black-box access, where only a small number of responses can be sampled for each query. In this setting, estimating the effective semantic alphabet size--that is, the number of distinct meanings expressed in the sampled responses--provides a useful proxy for downstream risk. However, frequency-based estimators tend to undercount rare semantic modes when the sample size is small, while graph-spectral quantities alone are not designed to estimate semantic occupancy accurately. To address this issue, we propose SHADE (Soft-Hybrid Alphabet Dynamic Estimator), a simple and interpretable estimator that combines Generalized Good-Turing coverage with a heat-kernel trace of the normalized Laplacian constructed from an entailment-weighted graph over sampled responses. The estimated coverage adaptively determines the fusion rule: under high coverage, SHADE uses a convex combination of the two signals, while under low coverage it applies a LogSumExp fusion to emphasize missing or weakly observed semantic modes. A finite-sample correction is then introduced to stabilize the resulting cardinality estimate before converting it into a coverage-adjusted semantic entropy score. Experiments on pooled semantic alphabet-size estimation against large-sample references and on QA incorrectness detection show that SHADE achieves the strongest improvements in the most sample-limited regime, while the performance gap narrows as the number of samples increases. These results suggest that hybrid semantic occupancy estimation is particularly beneficial when black-box uncertainty quantification must operate under tight sampling budgets.

cs.CL↗
arXiv2026

GatedCLIP: Gated Multimodal Fusion for Hateful Memes Detection

Detecting hateful content in multimodal memes presents unique challenges, as harmful messages often emerge from the complex interplay between benign images and text. We propose GatedCLIP, a Vision-Language model that enhances CLIP's multimodal capabilities with specialized architectural improvements for hateful memes detection. Our approach introduces learned projection heads that map CLIP embeddings to a task-optimized semantic space, a dynamic gated fusion mechanism that adaptively weights visual and textual features, and a contrastive learning objective that maintains cross-modal semantic alignment. Experiments on the Hateful Memes dataset demonstrate that GatedCLIP achieves an AUROC of 0.66, substantially outperforming the CLIP baseline (AUROC 0.49) while maintaining computational efficiency with only 350K trainable parameters.

cs.CV↗
arXiv2025

A Giant Peanut-shaped Ultra-High-Energy Gamma-Ray Emitter Off the Galactic Plane

Ultra-high-energy (UHE), exceeding 100 TeV (10^12 electronvolts), γ-rays manifests extreme particle acceleration in astrophysical sources. Recent observations by γ-ray telescopes, particularly by the Large High Altitude Air Shower Observatory (LHAASO), have revealed a few tens of UHE sources, indicating numerous Galactic sources capable of accelerating particles to PeV (10^15 electronvolts) energies. However, discerning the dominant acceleration mechanisms (leptonic versus hadronic), the relative contributions of specific source classes, and the role of particle transport in shaping their observed emission are central goals of modern UHE astrophysics. Here we report the discovery of a giant UHE γ-ray emitter at -17.5° off the Galactic plane - a region where UHE γ-ray sources are rarely found. The emitter exhibits a distinctive asymmetric shape, resembling a giant "Peanut" spanning 0.45° \times 4.6°, indicative of anisotropic particle distribution over a large area. A highly aged millisecond pulsar (MSP) J0218+4232 is the sole candidate accelerator positionally coincident with the Peanut region. Its association with UHE γ-rays extending to 0.7 PeV, if confirmed, would provide the first evidence of a millisecond pulsar powering PeV particles. Such a finding challenges prevailing models, which posit that millisecond pulsars cannot sustain acceleration to PeV energies. The detection reveals fundamental gaps in understanding particle acceleration, cosmic-ray transport, and interstellar magnetic field effects, potentially revealing new PeV accelerator (PeVatron) classes.

astro-ph.HE↗
arXiv2024

Ill-posedness and global solution for the $b$-equation

In this paper, we consider the Cauchy problem for the $b$-equation. Firstly, for $s>\frac32,$ if $u_{0}(x)\in H^{s}(\mathbb{R})$ and $m_{0}(x)=u_{0}(x)-u_{0xx}(x)\in L^{1}(\mathbb{R}),$ the global solutions of the $b$-equation is established when $b\geq1$ or $b\leq1.$ It's worth noting that our global result is a new result which doesn't need the condition that $m_{0}(x)$ keeps its sign. For $s<\frac32,$ it is shown (see [13]) that the Cauchy problem of the $b$-equation is ill-posed in Sobolev space $H^{s}(\mathbb{R})$ when $b>1$ or $b<1.$ In the present paper, for $s=\frac32,$ we prove that the Cauchy problem of the $b$-equation is also ill-posed in $H^{\frac32}(\mathbb{R})$ in the sense of norm inflation by constructing a class of special initial data when $b\neq1.$

math.AP↗
arXiv2022

A new result for the local well-posedness of the Camassa-Holm type equations in critial Besov spaces $B^{1+\frac{1}{p}}_{p,1},1\leq p<+\infty$

For the famous Camassa-Holm equation, the well-posedness in $B^{1+\frac{1}{p}}_{p,1}(\mathbb{R})$ with $1\leq p\leq2$ and the ill-posedness in $B^{1+\frac{1}{p}}_{p,r}(\mathbb{R})$ with $1\leq p\leq+\infty,\ 1<r\leq+\infty$ had been studied in \cite{d1,d2,glmy}. That is to say, it left an open problem in the critical case $B^{1+\frac{1}{p}}_{p,1}(\mathbb{R})$ with $2<p\leq+\infty$ proposed by Danchin in \cite{d1,d2}. In this paper, we solve this problem. The main difficulty is to prove the uniqueness, which usually needs to use the Moser-type inequality, resulting in the index $p$ belongs to $[1,2]$. To overcome the difficulty, inspired by Linares, Ponce and Thomas \cite{lps}, we combine the Lagrange coordinate transformation and small time conditions to avoid using the Moser-type inequality. As a result, we obtain the local well-posedness for the Camassa-Holm equation in critical Besov spaces $B^{1+\frac{1}{p}}_{p,1}(\mathbb{R})$ with $1\leq p<+\infty$. It is worth mentioning that our method is suitable for many Camassa-Holm type equations such as the Novikov equation and the two-component Camassa-Holm system, which can also improve their index on the local well-posedness.

math.AP↗
arXiv2022

Ill-posedness for the Cauchy problem of the Camassa-Holm equation in $B^{1}_{\infty,1}(\mathbb{R})$

For the famous Camassa-Holm equation, the well-posedness in $B^{1+\frac{1}{p}}_{p,1}(\mathbb{R})$ with $ p\in [1,\infty)$ and the ill-posedness in $B^{1+\frac{1}{p}}_{p,r}(\mathbb{R})$ with $ p\in [1,\infty],\ r\in (1,\infty]$ had been studied in \cite{d1,d2,glmy,yyg}, that is to say, it only left an open problem in the critical case $B^{1}_{\infty,1}(\mathbb{R})$ proposed by Danchin in \cite{d1,d2}. In this paper, we solve this problem by proving the norm inflation and hence the ill-posedness for the Camassa-Holm equation in $B^{1}_{\infty,1}(\mathbb{R})$. Therefore, the well-posedness and ill-posedness for the Camassa-Holm equation in all critial Besov spaces $B^{1+\frac{1}{p}}_{p,1}(\mathbb{R})$ with $ p\in [1,\infty]$ have been completed. Finally, since the norm inflation occurs by choosing an special initial data $u_0\in B^{1}_{\infty,1}(\mathbb{R})$ but $u^2_{0x}\notin B^{0}_{\infty,1}(\mathbb{R})$ (an example implies $B^{0}_{\infty,1}(\mathbb{R})$ is not a Banach algebra), we then prove that this condition is necessary. That is, if $u^2_{0x}\in B^{0}_{\infty,1}(\mathbb{R})$ holds, then the Camassa-Holm equation has a unique solution $u(t,x)\in \mathcal{C}_T(B^{1}_{\infty,1}(\mathbb{R}))\cap \mathcal{C}^{1}_T(B^{0}_{\infty,1}(\mathbb{R}))$ and the norm inflation will not occur.

math.AP↗
arXiv2022

A new result for the local well-posedness of the generalized Camassa-Holm equations in critial Besov spaces $B^{\frac{1}{p}}_{p,1},1\leq p<+\infty$

This paper is devoted to studying the local well-posedness (existence,uniqueness and continuous dependence) for the generalized Camassa-Holm equations in critial Besov spaces $B^{\frac{1}{p}}_{p,1}$ with $1\leq p<+\infty$, which improves the previous index $s> \max\{\frac{1}{2},\frac{1}{p}\}$ or $s=\frac{1}{p},\ p\in[1,2],\ r=1$ in \cite{linb,tu-yin4}. The main difficulty is to prove the uniqueness, which need to use the Moser-type inequality. To overcome the difficulty, we use the Lagrange coordinate transformation to obtain the uniqueness.

math.AP↗
arXiv2022

Ill-posedness for the Camassa-Holm equation in $B_{p,1}^{1}\cap C^{0,1}$

In this paper, we study the Cauchy problem for the Camassa-Holm equation on the real line. By presenting a new construction of initial data, we show that the solution map in the smaller space $B_{p,1}^{1}\cap C^{0,1}$ with $p\in(2,\infty]$ is discontinuous at origin. More precisely, $u_0\in B_{p,1}^{1}\cap C^{0,1}$ can guarantee that the Camassa-Holm equation has a unique local solution in $W^{1,p}\cap C^{0,1}$, however, this solution is instable and can have an inflation in $B_{p,1}^{1}\cap C^{0,1}$ for certain initial data.

math.AP↗
arXiv2021

The continuous dependence and non-uniform dependence of the rotation Camassa-Holm equation in Besov spaces

In this paper, we first establish the local well-posedness and continuous dependence for the rotation Camassa-Holm equation modelling the equatorial water waves with the weak Coriolis effect in nonhomogeneous Besov spaces $B^s_{p,r}$ with $s>1+1/p$ or $s=1+1/p,\ p\in[1,+\infty),\ r=1$ by a new way: the compactness argument and Lagrangian coordinate transformation, which removes the index constraint $s>3/2$ and improves our previous work \cite{guoy1}. Then, we prove the solution is not uniformly continuous dependence on the initial data in both supercritical and critical Besov spaces.

math.AP↗
arXiv2021

The well-posedness, ill-posedness and non-uniform dependence on initial data for the Fornberg-Whitham equation in Besov spaces

In this paper, we first establish the local well-posedness (existence, uniqueness and continuous dependence) for the Fornberg-Whitham equation in both supercritical Besov spaces $B^s_{p,r},\ s>1+\frac{1}{p},\ 1\leq p,r\leq+\infty$ and critical Besov spaces $B^{1+\frac{1}{p}}_{p,1},\ 1\leq p<+\infty$, which improves the previous work \cite{y2,ho,ht}. Then, we prove the solution is not uniformly continuous dependence on the initial data in supercritical Besov spaces $B^s_{p,r},\ s>1+\frac{1}{p},\ 1\leq p\leq+\infty,\ 1\leq r<+\infty$ and critical Besov spaces $B^{1+\frac{1}{p}}_{p,1},\ 1\leq p<+\infty$. At last, we show that the solution is ill-posed in $B^σ_{p,\infty}$ with $σ>3+\frac{1}{p},\ 1\leq p\leq+\infty$.

math.AP↗
arXiv2021

Ill-posedness for the higher dimensional Camassa-Holm equations in Besov spaces

In the paper, by constructing a initial data $u_{0}\in B^σ_{p,\infty}$ with $σ-2>\max\{1+\frac 1 p, \frac 3 2\}$, we prove that the corresponding solution to the higher dimensional Camassa-Holm equations starting from $u_{0}$ is discontinuous at $t=0$ in the norm of $B^σ_{p,\infty}$, which implies that the ill-posedness for the higher dimensional Camassa-Holm equations in $B^σ_{p,\infty}$.

math.AP↗
arXiv2020

The existence and uniqueness of global admissible conservative weak solution for the periodic single-cycle pulse equation

This paper is devoted to the study of the existence and uniqueness of global admissible conservative weak solutions for the periodic single-cycle pulse equation. We first transform the equation into an equivalent semilinear system by introducing a new set of variables. Using the standard ordinary differential equation theory, we then obtain the global solution to the semilinear system. Next, returning to the original coordinates, we get the global admissible conservative weak solution for the periodic single-cycle pulse equation. Finally, given an admissible conservative weak solution, we find a equation to single out a unique characteristic curve through each initial point and prove the uniqueness of global admissible conservative weak solution without any additional assumptions.

math.AP↗
arXiv2020

Prospects for a Multi-TeV Gamma-ray Sky Survey with the LHAASO Water Cherenkov Detector Array

The Water Cherenkov Detector Array (WCDA) is a major component of the Large High Altitude Air Shower Array Observatory (LHAASO), a new generation cosmic-ray experiment with unprecedented sensitivity, currently under construction. The WCDA is aimed at the study of TeV $γ$-rays. In order to evaluate the prospects of searching for TeV $γ$-ray sources with the WCDA, we present in this paper a projection for the one-year sensitivity of the WCDA to TeV $γ$-ray sources from TeVCat (footnote: http://tevcat.uchicago.edu) using an all-sky approach. Out of 128 TeVCat sources observable to the WCDA up to a zenith angle of $45^\circ$, we estimate that 42 would be detectable for one year of observations at a median energy of 1 TeV. Most of them are Galactic sources, and the extragalactic sources are Active Galactic Nuclei (AGN).

astro-ph.HE↗