Searcharxiv⌕ Search

arXiv subjects

Nicholas Meyer

Publications and source records attributed to Nicholas Meyer.

5 recordsLinked to original sources

An Obstruction to Realizing Aspherical $3$-Manifold Groups via Torus Surgery

Let $Y$ be a closed, connected, oriented, aspherical $3$-manifold with fundamental group $π= π_1(Y)$ and positive first Betti number $β_1(Y) > 0$, and let $X_0 = Y \times S^1$. We prove that no torus surgery on $X_0$ along a torus-fibered $T^2$-knot $T \subseteq X_0$ can yield a $4$-manifold $X = X_{0; T, f}$ with fundamental group $π$. The obstruction combines the vanishing of the Euler characteristic of $X$ with a classifying-map argument showing that the unimodular intersection form on $H^2(X; \mathbb{Z})$ must admit an isotropic subgroup of rank $β_1(π)$, forcing $χ(X) \geq 2$. Along the way, we also present an independent geometric proof via covering space theory, the Sphere Theorem, and Scott's Core Theorem that the fundamental group of a closed aspherical $3$-manifold is torsion-free.

math.GT↗

Extending Fibrations of the $3$-Torus and Applications to Torus Surgery in $4$-Manifolds

Suppose that $W$ and $W'$ are smooth, compact, and oriented $4$-manifolds that are either diffeomorphic to $S^1$ times the exterior $E_Y(K)$ of a fibered knot $K$ in a closed, connected, orientable $3$-manifold $Y$, or are diffeomorphic to $Σ_{g,1}$ bundles over the $2$-torus with monodromy fixing the boundary of the fiber pointwise. If $f: \partial W' \to \partial W$ is an orientation-preserving diffeomorphism of the $3$-torus boundaries, we have that $X = W \cup_f W'$ is a closed, oriented $4$-manifold that fibers over $S^1$. In particular, if $W' = T^2 \times D^2$ and $W = S^1 \times E_Y(K)$, then our result shows that the result of doing torus surgery in $S^1\times Y$ along $S^1 \times K$ is a $4$-manifold that fibers over $S^1$. Furthermore, we extend work of Zentner by showing that the result of torus surgery along $S^1$ times the unknot $\mathcal{U}$ in $S^1 \times S^3$ is diffeomorphic to $S^1$ times a lens space.

math.GT↗

VideoConviction: A Multimodal Benchmark for Human Conviction and Stock Market Recommendations

Social media has amplified the reach of financial influencers known as "finfluencers," who share stock recommendations on platforms like YouTube. Understanding their influence requires analyzing multimodal signals like tone, delivery style, and facial expressions, which extend beyond text-based financial analysis. We introduce VideoConviction, a multimodal dataset with 6,000+ expert annotations, produced through 457 hours of human effort, to benchmark multimodal large language models (MLLMs) and text-based large language models (LLMs) in financial discourse. Our results show that while multimodal inputs improve stock ticker extraction (e.g., extracting Apple's ticker AAPL), both MLLMs and LLMs struggle to distinguish investment actions and conviction--the strength of belief conveyed through confident delivery and detailed reasoning--often misclassifying general commentary as definitive recommendations. While high-conviction recommendations perform better than low-conviction ones, they still underperform the popular S\&P 500 index fund. An inverse strategy--betting against finfluencer recommendations--outperforms the S\&P 500 by 6.8\% in annual returns but carries greater risk (Sharpe ratio of 0.41 vs. 0.65). Our benchmark enables a diverse evaluation of multimodal tasks, comparing model performance on both full video and segmented video inputs. This enables deeper advancements in multimodal financial research. Our code, dataset, and evaluation leaderboard are available under the CC BY-NC 4.0 license.

cs.MM↗

DEFLOW: Self-supervised 3D Motion Estimation of Debris Flow

Existing work on scene flow estimation focuses on autonomous driving and mobile robotics, while automated solutions are lacking for motion in nature, such as that exhibited by debris flows. We propose DEFLOW, a model for 3D motion estimation of debris flows, together with a newly captured dataset. We adopt a novel multi-level sensor fusion architecture and self-supervision to incorporate the inductive biases of the scene. We further adopt a multi-frame temporal processing module to enable flow speed estimation over time. Our model achieves state-of-the-art optical flow and depth estimation on our dataset, and fully automates the motion estimation for debris flows. The source code and dataset are available at project page.

cs.CV↗

Tri-plane diagrams for simple surfaces in $S^4$

Meier and Zupan proved that an orientable surface $\mathcal{K}$ in $S^4$ admits a tri-plane diagram with zero crossings if and only if $\mathcal{K}$ is unknotted, so that the crossing number of $\mathcal{K}$ is zero. We determine the minimal crossing numbers of nonorientable unknotted surfaces in $S^4$, proving that $c(\mathcal{P}^{n,m}) = \max\{1,|n-m|\}$, where $\mathcal{P}^{n,m}$ denotes the connected sum of $n$ unknotted projective planes with normal Euler number $+2$ and $m$ unknotted projective planes with normal Euler number $-2$. In addition, we convert Yoshikawa's table of knotted surface ch-diagrams to tri-plane diagrams, finding the minimal bridge number for each surface in the table and providing upper bounds for the crossing numbers.

math.GT↗