arXiv subjects

Shipeng Wang

Publications and source records attributed to Shipeng Wang.

13 recordsLinked to original sources

arXiv2026

The homomorphism threshold of odd cycle $C_{2k-1}$ is below $\frac{1}{2k-1}$

The homomorphism threshold $δ_{\text{hom}}(H)$ of a graph $H$ asks how large the minimum degree of an $H$-free graph has to be in order to force a homomorphism to a bounded $H$-free graph. Determining this threshold is in general very difficult, and the odd cycles $C_{2k-1}$ are among the most important open cases for $k\ge 3$. Ebsen and Schacht proved the general upper bound $δ_{\text{hom}}(C_{2k-1})\le\frac{1}{2k-1}$, while a breakthrough of Sankar, using topological methods and a graph-theoretic analogue of homotopy equivalence, gave the first positive lower bound. The value $\frac{1}{2k-1}$ appeared particularly compelling: Ebsen and Schacht obtained the same exact threshold when all odd cycles of length at most $2k-1$ are forbidden, Huang, Liu, Rong and Xu later proved that it is the exact blowup threshold of $C_{2k-1}$, and Letzter and Snyder also explicitly asked whether $δ_{\text{hom}}(C_{5})=\frac{1}{5}$. Surprisingly, we show that the upper bound can be improved. More precisely, for every integer $k\ge3$, we prove $$ \frac{1}{2\left((k-1)^{4k-5}(2k-1)+\frac{(k-1)^{4k-5}-1}{k-2}\right)} \le δ_{\text{hom}}(C_{2k-1}) \le \frac{4(k-1)}{4(k-1)(2k-1)+1} < \frac{1}{2k-1}. $$ The new lower bound comes from a new graph-theoretic construction based on a sparse homomorphism theorem of Nešetřil and Zhu, and it improves Sankar's quantitative bound. The improved upper bound follows from a new structural argument that controls common neighborhoods along short odd paths. Our results have various consequences, in particular, every odd cycle of length at least five has pairwise distinct chromatic, homomorphism, polynomial removal, and linear removal thresholds, resolving two conjectures of Fox and Wigderson.

math.CO↗

arXiv2026

On the structure of dense graphs with given odd girth

A classical theorem of Andrásfai, Erdős, and Sós states that every $n$-vertex graph $G$ with odd girth at least $2k+1$ and minimum degree $δ(G)>\frac{2n}{2k+1}$ is bipartite (i.e., homomorphic to $K_2$). Messuti and Schacht proved that the same odd girth condition with $δ(G)>\frac{3n}{4k}$ forces a homomorphism to $C_{2k+1}$. In this paper, we strengthen the above results by showing that every $n$-vertex graph $G$ with odd girth at least $2k+1$ and minimum degree $δ(G)>\frac{4n}{6k-1}$ is homomorphic to the Möbius ladder on $4k$ vertices. This answers a question of Messuti and Schacht and generalizes a result of Brandt and Ribe-Baumann.

math.CO↗

arXiv2026

Towards Interpretable Federated Learning

Federated learning (FL) enables multiple data owners to build machine learning models collaboratively without exposing their private local data. In order for FL to achieve widespread adoption, it is important to balance the need for performance, privacy-preservation and interpretability, especially in mission critical applications such as finance and healthcare. Thus, interpretable federated learning (IFL) has become an emerging topic of research attracting significant interest from the academia and the industry alike. Its interdisciplinary nature can be challenging for new researchers to pick up. In this paper, we bridge this gap by providing (to the best of our knowledge) the first survey on IFL. We propose a unique IFL taxonomy which covers relevant works enabling FL models to explain the prediction results, support model debugging, and provide insights into the contributions made by individual data owners or data samples, which in turn, is crucial for allocating rewards fairly to motivate active and reliable participation in FL. We conduct comprehensive analysis of the representative IFL approaches, the commonly adopted performance evaluation metrics, and promising directions towards building versatile IFL techniques.

cs.LG↗

arXiv2025

Extremal triangle-free graphs with chromatic number at least four

Let $G$ be an $n$-vertex triangle-free graph. The celebrated Mantel's theorem showed that $e(G)\leq \lfloor\frac{n^2}{4}\rfloor$. In 1962, Erdős (together with Gallai), and independently Andrásfai, proved that if $G$ is non-bipartite then $e(G)\leq \lfloor\frac{(n-1)^2}{4}\rfloor+1$. In this paper, we extend this result and show that if $G$ has chromatic number at least four and $n\geq 90$, then $e(G)\leq \lfloor\frac{(n-3)^2}{4}\rfloor+5$. The blow-ups of Grötzsch graph shows that this bound is best possible.

math.CO↗

arXiv2025

Longest odd cycles in non-bipartite $C_{2k+1}$-free graphs

In strengthening a result of Andrásfai, Erdős and Sós in 1974, Häggkvist proved that if $G$ is an $n$-vertex $C_{2k+1}$-free graph with minimum degree $δ(G)>\frac{2n}{2k+3}$ and $n>\binom{k+2}{2}(2k+3)(3k+2)$, then $G$ contains no odd cycle of length greater than $\frac{k+1}{2}$. This result has many applications.In this paper, we consider a similar problem by replacing minimum degree condition with edge number condition. We prove that for integers $n,k,r$ with $k\geq 2,3\leq r\leq 2k$ and $n \geq 2\left(r+2\right)\left(r+1\right)\left(r+2k\right)$, if $G$ is an $n$-vertex $C_{2k+1}$-free graph with $e(G) \geq \left\lfloor\frac{(n-r+1)^2}{4}\right\rfloor+\binom{r}{2}$, then $G$ contains no odd cycle of length greater than $r$. The construction shows that the result is best possible. This extends a result of Brandt [Discrete Applied Mathematics 79 (1997)], and a result of Bollobás and Thomason [Journal of Combinatorial Theory, Series B. 77 (1999)], and a result of Caccetta and Jia [Graphs Combin. 18 (2002)] and independently proving by Lin, Ning and Wu [Combin. Probab. Comput. 30 (2021)]. Recently, Ren, Wang, Yang, and the second author [SIAM J. Discrete Math. 38 (2024)] show that for $3\leq r\leq 2k$ and $n\geq 318(r-2)^2k$, every $n$-vertex $C_{2k+1}$-free graph with $e(G) \geq \left\lfloor\frac{(n-r+1)^2}{4}\right\rfloor+\binom{r}{2}$ can be made bipartite by deleting at most $r-2$ vertices or deleting at most $\binom{\lfloor\frac{r}{2}\rfloor}{2}+\binom{\lceil\frac{r}{2}\rceil}{2}$ edges. As an application, we derive this result and provide a simple proof.

math.CO↗

arXiv2025

Characterizing forbidden pairs for spanning $\varTheta$-subgraphs of 2-connected graphs

Let $\mathcal{F}$ be a set of connected graphs, and let $G$ be a graph. We say that $G$ is \emph{$\mathcal{F}$-free} if it does not contain $F$ as an induced subgraph for all $F\in\mathcal{F}$, and we call $\mathcal{F}$ a forbidden pair if $|\mathcal{F}|=2$. A \emph{$\varTheta$-graph} is the graph consisting of three internally disjoint paths with the same pair of end-vertices. If the $\varTheta$-subgraph $T$ contains all vertices of $G$, then we call $T$ a \emph{spanning $\varTheta$-subgraph} of $G$. In this paper, we characterize all pairs of connected graphs $R,S$ such that every 2-connected $\{R,S\}$-free graph has a spanning $\varTheta$-subgraph. In order to obtain this result, we also characterize all minimal 2-connected non-cycle claw-free graphs without spanning $\varTheta$-subgraphs.

math.CO↗

arXiv2023

Towards AI-Empowered Crowdsourcing

Crowdsourcing, in which human intelligence and productivity is dynamically mobilized to tackle tasks too complex for automation alone to handle, has grown to be an important research topic and inspired new businesses (e.g., Uber, Airbnb). Over the years, crowdsourcing has morphed from providing a platform where workers and tasks can be matched up manually into one which leverages data-driven algorithmic management approaches powered by artificial intelligence (AI) to achieve increasingly sophisticated optimization objectives. In this paper, we provide a survey presenting a unique systematic overview on how AI can empower crowdsourcing to improve its efficiency - which we refer to as AI-Empowered Crowdsourcing(AIEC). We propose a taxonomy which divides AIEC into three major areas: 1) task delegation, 2) motivating workers, and 3) quality control, focusing on the major objectives which need to be accomplished. We discuss the limitations and insights, and curate the challenges of doing research in each of these areas to highlight promising future research directions.

cs.CY↗

arXiv2023

A stability result for $C_{2k+1}$-free graphs

A graph $G$ is called $C_{2k+1}$-free if it does not contain any cycle of length $2k+1$. In 1981, Haggkvist, Faudree and Schelp showed that every $n$-vertex triangle-free graph with more than $\frac{(n-1)^2}{4}+1$ edges is bipartite. In this paper, we extend their result and show that for $1\leq t\leq 2k-2$ and $n\geq 318t^2k$, every $n$-vertex $C_{2k+1}$-free graph with more than $\frac{(n-t-1)^2}{4}+\binom{t+2}{2}$ edges can be made bipartite by either deleting at most $t-1$ vertices or deleting at most $\binom{\lfloor\frac{t+2}{2}\rfloor}{2}+\binom{\lceil\frac{t+2}{2}\rceil}{2}-1$ edges. The construction shows that this is best possible.

math.CO↗

arXiv2023

Interactive Molecular Discovery with Natural Language

Natural language is expected to be a key medium for various human-machine interactions in the era of large language models. When it comes to the biochemistry field, a series of tasks around molecules (e.g., property prediction, molecule mining, etc.) are of great significance while having a high technical threshold. Bridging the molecule expressions in natural language and chemical language can not only hugely improve the interpretability and reduce the operation difficulty of these tasks, but also fuse the chemical knowledge scattered in complementary materials for a deeper comprehension of molecules. Based on these benefits, we propose the conversational molecular design, a novel task adopting natural language for describing and editing target molecules. To better accomplish this task, we design ChatMol, a knowledgeable and versatile generative pre-trained model, enhanced by injecting experimental property information, molecular spatial knowledge, and the associations between natural and chemical languages into it. Several typical solutions including large language models (e.g., ChatGPT) are evaluated, proving the challenge of conversational molecular design and the effectiveness of our knowledge enhancement method. Case observations and analysis are conducted to provide directions for further exploration of natural-language interaction in molecular discovery.

cs.CL↗

arXiv2022

Counterexamples to Gerbner's Conjecture on Stability of Maximal $F$-free Graphs

Let $F$ be an $(r+1)$-color critical graph with $r\geq 2$, that is, $χ(F)=r+1$ and there is an edge $e$ in $F$ such that $χ(F-e)=r$. Gerbner recently conjectured that every $n$-vertex maximal $F$-free graph with at least $(1-\frac{1}{r})\frac{n^2}{2}- o(n^{\frac{r+1}{r}})$ edges contains an induced complete $r$-partite graph on $n-o(n)$ vertices. Let $F_{s,k}$ be a graph obtained from $s$ copies of $C_{2k+1}$ by sharing a common edge. In this paper, we show that for all $k\geq 2$ if $G$ is an $n$-vertex maximal $F_{s,k}$-free graph with at least $n^{2}/4 - o(n^{\frac{s+2}{s+1}})$ edges, then $G$ contains an induced complete bipartite graph on $n-o(n)$ vertices. We also show that it is best possible. This disproves Gerbner's conjecture for $r=2$.

math.CO↗

arXiv2021

A Stability Theorem for Maximal $C_{2k+1}$-free Graphs

For any positive integer $k$, we show that every maximal $C_{2k+1}$-free graph with at least $n^2/4-o(n^{3/2})$ edges contains an induced complete bipartite subgraph on $(1-o(1))n$ vertices. We also show that this is best possible.

math.CO↗

arXiv2021

Training Networks in Null Space of Feature Covariance for Continual Learning

In the setting of continual learning, a network is trained on a sequence of tasks, and suffers from catastrophic forgetting. To balance plasticity and stability of network in continual learning, in this paper, we propose a novel network training algorithm called Adam-NSCL, which sequentially optimizes network parameters in the null space of previous tasks. We first propose two mathematical conditions respectively for achieving network stability and plasticity in continual learning. Based on them, the network training for sequential tasks can be simply achieved by projecting the candidate parameter update into the approximate null space of all previous tasks in the network training process, where the candidate parameter update can be generated by Adam. The approximate null space can be derived by applying singular value decomposition to the uncentered covariance matrix of all input features of previous tasks for each linear layer. For efficiency, the uncentered covariance matrix can be incrementally computed after learning each task. We also empirically verify the rationality of the approximate null space at each linear layer. We apply our approach to training networks for continual learning on benchmark datasets of CIFAR-100 and TinyImageNet, and the results suggest that the proposed approach outperforms or matches the state-ot-the-art continual learning approaches.

cs.LG↗

arXiv2018

HyperAdam: A Learnable Task-Adaptive Adam for Network Training

Deep neural networks are traditionally trained using human-designed stochastic optimization algorithms, such as SGD and Adam. Recently, the approach of learning to optimize network parameters has emerged as a promising research topic. However, these learned black-box optimizers sometimes do not fully utilize the experience in human-designed optimizers, therefore have limitation in generalization ability. In this paper, a new optimizer, dubbed as \textit{HyperAdam}, is proposed that combines the idea of "learning to optimize" and traditional Adam optimizer. Given a network for training, its parameter update in each iteration generated by HyperAdam is an adaptive combination of multiple updates generated by Adam with varying decay rates. The combination weights and decay rates in HyperAdam are adaptively learned depending on the task. HyperAdam is modeled as a recurrent neural network with AdamCell, WeightCell and StateCell. It is justified to be state-of-the-art for various network training, such as multilayer perceptron, CNN and LSTM.

cs.LG↗