arXiv++ Combinatorics

Forty-five years ago, a young researcher in finite permutation group theory encountered a paper by Robert Woodrow. The homogeneous triangle-free graph Woodrow described there seemed to be an infinite analogue of the Higman--Sims graph which had played an important role in the researcher's thesis. The encounter changed the course of the researcher's career. This paper is the story of that event and its aftermath. The final section of the paper suggests that Fraïssé classes of rigid structures are a potentially interesting generalisation of Ramsey classes.

We extend our previous algorithm that generates all labeled graphs with a given graphical degree sequence to generate all labeled triangle-free graphs with a given graphical degree sequence. The algorithm uses various pruning techniques to avoid having to first generate all labeled realizations of the input sequence and then testing whether each labeled realization is triangle-free. It can be further extended to generate all labeled bipartite graphs with a given graphical degree sequence by adding a simple test whether each generated triangle-free realization is a bipartite graph. All output graphs are generated in the lexicographical ordering as in the original algorithm. The algorithms can also be easily parallelized.

The classic upper bound on the chromatic number of $d$-degenerate graphs is $d+1$, shown to be tight by complete graphs. A natural question is whether this bound remains tight if one forbids large cliques. Classic constructions of Tutte and Zykov from the early 50s show that there exist $d$-degenerate $(d+1)$-chromatic graphs that are triangle-free, however these constructions grow rapidly with $d$. Motivated by this and addressing a problem posed by the second author at the Oberwolfach Graph Theory workshop, we prove that the minimum order $f(d)$ of a $d$-degenerate triangle-free graph of chromatic number $d+1$ satisfies $e^{Ω(d)}\le f(d)\le e^{O(d^2\log d)}.$ The lower bound follows from a novel upper bound on the chromatic number of triangle-free graphs: Every triangle-free $d$-degenerate graph $G$ on $n \le e^{O(d)}$ vertices satisfies $$χ(G)\le O\left(\frac{d}{\log\left(d/\log n\right)}\right).$$ We extend this to a more general result about degenerate graphs with sparse neighborhoods, which has applications to many graph coloring problems: For example, we prove that every counterexample to Hadwiger's conjecture with parameter $t$ must have a complete bipartite subgraph with one exponentially large side ($K_{a,b}$ where $a=(\log t)^{1/2-o(1)}$ and $b=e^{t^{1-o(1)}}$) or a small and very dense subgraph (of order $\le t$ with $t^{2-o(1)}$ edges) in some neighborhood. For the upper bound on $f(d)$ we establish a surprising connection between $f(d)$ and the on-line-chromatic number $g(n)$ of $n$-vertex triangle-free graphs. We also give an asymptotic improvement of the previous best upper bound for $g(n)$ due to Lovász, Saks and Trotter from 1989. Along the way we disprove a generalization of Harris' fractional coloring conjecture to graphs of bounded clique number and raise numerous problems which open up interesting directions to explore for future research.

Given a triangle-free graph $G$ with chromatic number $k$ and a proper vertex coloring $φ$ of $G$, it is conjectured that $G$ contains an induced rainbow path on $k$ vertices under $φ$. Scott and Seymour proved the existence of an induced rainbow path on $(\log \log \log k)^{\frac{1}{3}- o(1)}$ vertices. We improve this to $(\log k)^{\frac{1}{2}- o(1)}$ vertices. Further, we prove the existence of an induced path that sees $\frac{k}{2}$ colors.

The chromatic number $χ$ of a graph is bounded from below by its clique number $ω,$ but it can be arbitrary large. Perfect graphs are defined by $χ=ω$ for all induced subgraphs. An interesting relaxation are $χ$-bounded graph classes, where $χ\leq f(ω).$ It is not always possible to achieve this with a polynomial $f.$ The edge ideal $I_G$ of a graph $G$ is generated by monomials $x_ux_v$ for each edge $uv$ of $G.$ The bi-graded betti numbers $β_{i,j}(I)$ are central algebraic geometric invariants. We study the graph classes where for some fixed $i,j$ that syzygy vanishes, that is, $β_{i,j}(I_G)=0.$ We prove that $χ\leq f(ω),$ where $f$ is a polynomial of degree $2j-2i-4.$ For the elementary special case $β_{i,2i+2}(I_G)=0,$ this amounts to that $(i+1)K_2$-free graphs are ${ω-1+2i \choose 2i}$-colorable, improving on an old combinatorial result by Wagon. We also show that triangle-free graphs with $β_{i,j}(I_G)=0$ are $(j-1)$-colorable. Complexity wise, we show that these colorings can be derived in time $O(n^3)$ for graphs on $n$ vertices. Moreover, we show that for almost all graphs with parabolic $i,j,$ there are better bounds on $χ.$

A graph $G$ is $k$-vertex-critical if $χ(G)=k$, but $χ(G')<k$ for every proper induced subgraph $G'$ of $G$. For a family of graphs $\mathcal{F}$, $G$ is $\mathcal{F}$-free if no graph $F \in \mathcal{F}$ is an induced subgraph of $G$. We show that there are exactly three 4-vertex-critical $\{P_7,C_3\}$-free graphs containing an induced $C_7$, thereby settling the first of the two cases of a conjecture by Goedgebeur and Schaudt [J.~Graph Theory, 87:188--207, 2018]. Moreover, we show that all $\{P_5+P_1,C_3\}$-free graphs are $3$-colorable and by combining our result with known results from the literature, we completely characterize the maximum chromatic number of $\{F,C_3\}$-free graphs if $F$ is a six-vertex induced subgraph of $P_7$. Finally, we construct an infinite family of $4$-vertex-critical $\{4K_2,C_3\}$-free graphs. These graphs are also $\{P_{11},C_3\}$-free and this is the first value of $t$ for which an infinite family of $4$-vertex-critical $\{P_{t},C_3\}$-free graphs is known.

In this paper, we characterize graphs with circular chromatic number less than 3 in terms of certain balancing labellings studied in the context of signed graphs. In fact, we construct a signed graph which is universal for all such labellings of graphs with circular chromatic number less than $3$, and is closely related to the generic circular triangle-free graph studied by Bodirsky and Guzmán-Pro. Moreover, our universal structure gives rise to a representation of the relation algebra $56_{65}$. We then use this representation to show that the network satisfaction problem described by this relation algebra belongs to NP. This concludes the full classification of the existence of a universal square representation, as well as the complexity of the corresponding network satisfaction problem, for relation algebras with at most four atoms.

Alon, Krivelevich, and Sudakov conjectured in 1999 that every $F$-free graph of maximum degree at most $Δ$ has chromatic number $O(Δ/ \log Δ)$. This was previously known only for almost bipartite graphs, that is, for subgraphs of $K_{1,t,t}$ (verified by Alon, Krivelevich, and Sudakov themselves), while most recent results were concerned with improving the leading constant factor in the case where $F$ is almost bipartite. We prove this conjecture for all $3$-colorable graphs $F$, i.e. subgraphs of $K_{t,t,t}$, representing the first progress toward the conjecture since it was posed. A closely related conjecture of Ajtai, Erdős, Komlós, and Szemerédi from 1981 asserts that for every graph $F$, every $n$-vertex $F$-free graph of average degree $d$ contains an independent set of size $Ω(n \log d / d)$. We prove this conjecture in a strong form for all 3-colorable graphs $F$. More precisely, we show that every $n$-vertex $K_{t,t,t}$-free graph of average degree $d$ contains an independent set of size at least $(1 - o(1)) n \log d / d$, matching Shearer's celebrated bound for triangle-free graphs (the case $t = 1$) and thereby yielding a substantial strengthening of it. Our proof combines a new variant of the Rödl nibble method for constructing independent sets with a Turán-type result on $K_{t,t,t}$-free graphs.

In 2012, Mader conjectured that for any tree $T$ of order $m$, every $k$-connected graph $G$ with minimum degree at least $\lfloor \frac{3k}{2}\rfloor+m-1$ contains a subtree $T'\cong T$ such that $G-V(T')$ remains $k$-connected. In 2022, Luo, Tian, and Wu considered an analogous problem for bipartite graphs and conjectured that for any tree $T$ with bipartition $(X,Y)$, every $k$-connected bipartite graph $G$ with minimum degree at least $k+\max\{|X|,|Y|\}$ contains a subtree $T'\cong T$ such that $G-V(T')$ remains $k$-connected. In this paper, we relax the bipartite assumption by considering triangle-free graphs and prove that for any tree $T$ of order $m$, every $k$-connected triangle-free graph $G$ with minimum degree at least $2k+3m-4$ contains a subtree $T' \cong T$ such that $G-V(T')$ remains $k$-connected. Furthermore, we establish refined results for specific subclasses such as bipartite graphs or graphs with girth at least five.

A graph class is $χ$-bounded if the only way to force large chromatic number in graphs from the class is by forming a large clique. In the 1970s, Erdős conjectured that intersection graphs of straight-line segments in the plane are $χ$-bounded, but this was disproved by Pawlik et al. (2014), who showed another way to force large chromatic number in this class -- by triangle-free graphs $B_k$ with $χ(B_k)=k$ constructed by Burling (1965). This also disproved the celebrated conjecture of Scott (1997) that classes of graphs excluding induced subdivisions of a fixed graph are $χ$-bounded. We prove that in broad classes of graphs excluding induced subdivisions of a fixed graph, including the increasingly more general classes of segment intersection graphs, string graphs, region intersection graphs, and hereditary classes of graphs with finite asymptotic dimension, large chromatic number can be forced only by large cliques or large graphs $B_k$. One corollary is that the hereditary closure of $\{B_k\colon k\geq 1\}$ forms a minimal hereditary graph class with unbounded chromatic number -- the second known graph class with this property after the class of complete graphs. Another corollary is that the decision variant of approximate coloring in the aforementioned graph classes can be solved in polynomial time by exhaustively searching for a sufficiently large clique or copy of $B_k$. We also discuss how our results along with some results of Chudnovsky, Scott, and Seymour on the existence of colorings can be turned into polynomial-time algorithms for the search variant of approximate coloring in string graphs (with intersection model in the input) and other aforementioned graph classes. Such an algorithm has not yet been known for any graph class that is not $χ$-bounded.

Let $Q$ be an $s$-vertex $r$-uniform hypergraph, and let $H$ be an $n$-vertex $r$-uniform hypergraph. Denote by $\mathcal{N}(Q,H)$ the number of isomorphic copies of $Q$ in $H$. For a hereditary family $\mathcal{P}$ of $r$-uniform hypergraphs, define $$π(Q,\mathcal{P}):=\lim\limits_{n\to \infty}\binom{n}{s}^{-1}\max\{\mathcal{N}(Q,H): H\in \mathcal{P}~~\mbox{and}~~|V(H)|=n\}.$$ For $p\geq1$, the $(p,Q)$-spectral radius of $H$ is defined as $$λ^{(p)}(Q,H):=\max_{\|\mathbf{x}\|_{p}=1}s!\sum_{\{i_{1},\ldots,i_{s}\}\in \binom{[n]}{s}}\mathcal{N}(Q,H[\{i_{1},\ldots,i_{s}\}])x_{i_{1}}\cdots x_{i_{s}}.$$ %generalizing the concept of the $p$-spectral radius introduced by %Keevash, Lenz, and Mubayi \cite{KLM2014}. In this paper, we present a systematically investigation of the parameter $λ^{(p)}(Q,H)$. First, we prove that the limit $$λ^{(p)}(Q,\mathcal{P}):=\lim\limits_{n\to \infty}n^{s/p-s}\max\{λ^{(p)}(Q,H): H\in \mathcal{P}~~\mbox{and}~~|V(H)|=n\}$$ exists, and for $p>1$, it satisfies $$π(Q,\mathcal{P})=λ^{(p)}(Q,\mathcal{P}).$$ Second, we study spectral generalized Turán problems. Specifically, we establish a spectral stability result and apply it to derive a spectral version of the Erdős Pentagon Problem: for $p\geq1$ and sufficiently large $n$, the balanced blow-up of $C_{5}$ maximizes $λ^{(p)}(C_{5},H)$ among all $n$-vertex triangle-free graphs $H$, thereby improving a result of Liu \cite{Liu2025}. Furthermore, we show that for $p\geq1$ and sufficiently large $n$, the $l$-partite Turán graph $T_{l}(n)$ attains the maximum $λ^{(p)}(K_{s},H)$ among all $n$-vertex F-free graphs $H$, where $F$ is an edge-critical graph with $χ(F)=l+1$. This provides a spectral analogue of a theorem due to Ma and Qiu \cite{MQ2020}.

An odd independent set $S$ in a graph $G=(V,E)$ is an independent set of vertices such that, for every vertex $v \in V \setminus S$, either $N(v) \cap S = \emptyset$ or $|N(v) \cap S| \equiv 1$ (mod 2), where $N(v)$ stands for the open neighborhood of $v$. The largest cardinality of odd independent sets of a graph $G$, denoted $α_{od}(G)$, is called the odd independence number of $G$. This new parameter is a natural companion to the recently introduced strong odd chromatic number. A proper vertex coloring of a graph $G$ is a strong odd coloring if, for every vertex $v \in V(G)$, each color used in the neighborhood of $v$ appears an odd number of times in $N(v)$. The minimum number of colors in a strong odd coloring of $G$ is denoted by $χ_{so}(G)$. A simple relation involving these two parameters and the order $|G|$ of $G$ is $α_{od}(G)\cdot χ_{so}(G) \geq |G|$, parallel to the same on chromatic number and independence number. We develop several basic inequalities concerning $α_{od}(G)$, and use already existing results on strong odd coloring, to derive lower bounds for odd independence in many families of graphs. We prove that $α_{od}(G) = α(G^2)$ holds for all claw-free graphs $G$, and present many results, using various techniques, concerning the odd independence number of cycles, paths, Moore graphs, Kneser graphs, the complete subdivision $S(K_n)$ of $K_n$, the half graphs $H_{n,n}$, and $K_p \Box K_q$. Further, we consider the odd independence number of the hypercube $Q_d$ and also of the complements of triangle-free graphs. Many open problems for future research are stated.

Cho and van Willigenburg (arXiv:1508.07670) and Alinaeifard, Wang, and van Willgenburg (arXiv:2010.00147) introduce multiplicative chromatic bases for the ring $Λ$ of symmetric functions, consisting of the chromatic symmetric functions (CSFs) of a sequence of connected graphs $G_1,G_2,\dots$ such that $G_n$ has total weight $n$, together with the CSFs of their disjoint unions. In arXiv:1707.04058, Tsujie introduces an alternative ring structure $\widetildeΛ$ on the vector space $Λ$ that makes CSFs multiply over joins instead of over disjoint unions. The $\widetilde{m}_λ$ basis, consisting of all CSFs of weighted cliques, is a multiplicative basis for $\widetildeΛ$, as is the $r_λ$ basis of complete multipartite graphs studied by Penaguiao (arXiv:1803.08824) and Crew and Spirkl (arXiv:2009.14141). We show that one can get more of these "cochromatic bases" (where the starting graphs are combined by joins instead of disjoint unions, hence forming a multiplicative basis for $\widetildeΛ$ instead of $Λ$) if and only if the starting graphs are edgeless. We also show that $\widetildeΛ$ is a Hopf algebra with the same coproduct as $Λ$, and that many of the chromatic bases for $Λ$ generated by cliques can be taken to their corresponding cochromatic bases via Hopf algebra isomorphisms $Λ\to \widetildeΛ.$ We also show that there is a single Hopf algebra morphism taking the CSFs of all unweighted triangle-free graphs to the CSFs of their complements, and we give several more conditions and examples for when one can or cannot find Hopf algebra maps taking the CSFs of certain graphs to the CSFs of their complements. Finally, we show that $K$-analogues of many of the above statements also hold if one instead uses the Kromatic symmetric function (KSF) defined by Crew, Pechenik, and Spirkl (arXiv:2301.02177).

We study the chromatic number of typical triangle-free graphs with $Θ\left( n^{3/2} (\log n)^{1/2} \right)$ edges and establish the width of the scaling window for the transitions from $χ= 3$ to $χ= 4$ and from $χ= 4$ to $χ= 5$. The transition from $3$- to $4$-colorability has scaling window of width $Θ(n^{4/3} (\log n)^{-1/3})$. To prove this, we show a high probability equivalence of the $3$-colorability of a random triangle-free graph at this density and the satisfiability of an instance of bipartite random $2$-SAT, for which we establish the width of the scaling window following the techniques of Bollob{á}s, Borgs, Chayes, Kim, and Wilson. The transition from $4$- to $5$-colorability has scaling window of width $Θ(n^{3/2} (\log n)^{-1/2})$. To prove this, we show a high probability equivalence of the $4$-colorability of a random triangle-free graph at this density and the simultaneous $2$-colorability of two independent Erdős--Rényi random graphs. For this transition, we also establish the limiting probability of $4$-colorability inside the scaling window.

In the first part of this paper we determine the maximum size of a (finite, simple, connected) bipartite graph of given order, diameter $d$, and connectivity $κ$. It was shown by Ali, Mazorodze, Mukwembi and Vetrík [On size, order, diameter and edge-connectivity of graphs. Acta Math. Hungar. {\bf 152}, (2017)] that for a connected triangle-free graph of order $n$, diameter $d$ and edge-connectivity $λ$, the size is bounded from above by about $\frac{1}{4}\left(n-\frac{(λ+c) d}{2}\right)^2+O(n)$, where $c\in\{0, \frac{1}{3}, 1\}$ for different values of $λ$. In the second part of this paper we show that this bound by Ali et al. on the size can be improved significantly for a much larger subclass of triangle-free graphs, namely, bipartite graphs of order $n$, diameter $d$ and edge-connectivity $λ$. We prove our result only for $λ= 2, 3, 4$ because it can be observed from this paper by Ali et al. that for $λ\geq 5$, there exists $\ell$-edge-connected bipartite graphs of given order and diameter whose size differs from the maximal size for given minimum degree $\ell$ only by at most a constant. Also, unlike the approach in the proof on the size of triangle-free graphs by Ali et al., our proof employs a completely different technique, which enables us to identify the extremal graphs; hence the bounds presented here are sharp.

Generalized Sierpiński graphs constitute a distinctive class of fractal-like networks, whose self-similar properties have attracted growing attention. In particular, a number of graph invariants have been studied in generalized Sierpiński graphs. In this paper, we focus on injective colorings of this class of graphs, both the vertex and the edge version. The vertex version of injective colorings in generalized Sierpiński graphs was initiated in [Injective colorings of Sierpiński-like graphs and Kneser graphs,Graphs.\ Combin.\ 41 (2025) 83], where the authors determined the injective chromatic numbers of the standard Sierpiński graphs (which are those whose base graph is a clique) and asked about the values when the base graph is a cycle. We resolve this question by proving that $χ_i(S_{C_k}^n)=3$ for every $n\ge 2$ and every $k\ge 3$, which follows from a more general result on generalized Sierpi\' nski graphs $S_G^n$ with arbitrary base graphs $G$. Moreover, we prove (an almost conclusive result) that $χ_i(S_G^n)\in \{χ_i(G),χ_i(G)+1\}$ for any graph $G$ and any $n\ge 2$. Injective edge colorings appear to be more difficult, especially in graphs with triangles. On a positive note, we prove that $χ_i'(S_{K_3}^n)=5$ for all $n\ge 3$. Furthermore, if $G$ is a triangle-free graph, we prove that $χ_i'(S_G^n)\in \{χ_i'(S_G^3),χ_i'(S_G^3)+1\}$ for all $n\ge 4$, and provide some sufficient conditions on an injective edge coloring of the 3-dimensional Sierpiński graph over $G$, which ensure that $χ_i'(S_G^n)=χ_i'(S_G^3)$. In particular, the latter result enables us to establish that $χ_i'(S_{C_4}^n)=3$, $χ_i'(S_{C_5}^n)=4$ and $χ_i'(S_{C_6}^n)=3$ hold for any $n\ge 2$.

Let $\mathrm{ex}(n, H)$ be the Turán number of $H$ for a given graph $H$. A graph is color-critical if it contains an edge whose removal reduces its chromatic number. Simonovits' chromatic critical edge theorem states that if $H$ is color-critical with $χ(H)=k+1$, then there exists an $n_0(H)$ such that ex$(n, H)=e(T_{n,k})$ and the Turán graph $T_{n,k}$ is the only extremal graph provided $n\geq n_0(H).$ A book graph $B_{r+1}$ is a set of $r+1$ triangles with a common edge, where $r\geq0$ is an integer. Note that $B_{r+1}$ is a color-critical graph with $χ(B_{r+1})=3$. Simonovits' theorem implies that $T_{n,2}$ is the only extremal graph for $B_{r+1}$-free graphs of sufficiently large order $n$. Furthermore, Edwards and independently Khadžiivanov and Nikiforov completely confirmed Erdős' booksize conjecture and obtained that ex$(n, B_{r+1})=e(T_{n,2})$ for $n\geq n_0(B_{r+1})=6r$. Recently, Zhai and Lin [J. Graph Theory 102 (2023) 502-520] investigated the problem of booksize from a spectral perspective. Note that the extremal graph $T_{n,2}$ is bipartite. Motivated by the above elegant results, we in this paper focus on the Turán problem of non-bipartite $B_{r+1}$-free graphs of order $n$. For $r = 0,$ Erdős proved a nice result: If $G$ is a non-bipartite triangle-free graph on $n$ vertices, then $e(G)\leq\big\lfloor\frac{(n-1)^{2}}{4}\big\rfloor+1$. For general $r\geq1,$ we determine the exact value of Turán number of $B_{r+1}$ in non-bipartite graphs and characterize all extremal graphs provided $n$ is sufficiently large. An interesting phenomenon is that the Turán numbers and extremal graphs are completely different for $r=0$ and general $r\geq1.$

Mader conjectured that an average degree of at least $3k-1$ is sufficient for the existence of a $(k+1)$-connected subgraph. The following minimum degree version holds: Every graph with minimum degree at least $3k-1$ has a $(k+1)$-connected subgraph on more than $2k$ vertices. Moreover, for triangle-free graphs, already an average degree of at least $2k$ is sufficient for a $(k+1)$-connected subgraph (which has at least $2(k+1)$ vertices). For edge-connectedness (in simple graphs), we prove the following: Every graph of average degree at least $2k$ has a $(k+1)$-edge-connected subgraph on more than $2k$ vertices. Moreover, for every small $α>0$ and for $k$ large enough in terms of $α$, already a minimum degree of at least $k+k^{\frac{1}{2}+α} = \big(1+o(1)\big)k$ is sufficient for a $(k+1)$-edge-connected subgraph. It is shown that all of these results are sharp in some sense. The results can be used for the decomposition of graphs into two highly (edge-)connected parts.

The chromatic discrepancy of a graph $G$, denoted $φ(G)$, is the least over all proper colourings $σ$ of $G$ of the greatest difference between the number of colours $|σ(V(H))|$ spanned by an induced subgraph $H$ of $G$ and its chromatic number $χ(H)$. We prove that the chromatic discrepancy of a triangle-free graph $G$ is at least $χ(G)-2$. This is best possible and positively answers a question raised by Aravind, Kalyanasundaram, Sandeep, and Sivadasan. More generally, we say that a graph $G$ is locally $s$-colourable if the closed neighbourhood of any vertex $v\in V(G)$ is properly $s$-colourable; in particular, a triangle-free graph is locally $2$-colourable. We conjecture that every locally $s$-colourable graph $G$ satisfies $φ(G) \geq χ(G)-s$, and show that this would be almost best possible. We prove the conjecture when $χ(G) \le 11s/6$, and as a partial result towards the general case, we prove that every locally $s$-colourable graph $G$ satisfies $φ(G) \geq χ(G) - s\ln χ(G)$. If the conjecture holds, it implies in particular, for every integer $\ell\geq 2$, that any graph $G$ without any copy of $C_{\ell+1}$, the cycle of length $\ell+1$, satisfies $φ(G) \geq χ(G) - \ell$. When $\ell \ge 3$ and $G\neq K_\ell$, we conjecture that we actually have $φ(G)\ge χ(G) - \ell + 1$, and prove it in the special case $\ell = 3$ or $χ(G) \le 5\ell/3$. In general, we further obtain that every $C_{\ell+1}$-free graph $G$ satisfies $φ(G) \geq χ(G) - O_{\ell}(\ln \ln χ(G))$. We do so by determining an almost tight bound on the chromatic number of balls of radius at most $\ell/2$ in $G$, which could be of independent interest.

A well-known result of Mantel asserts that every $n$-vertex triangle-free graph $G$ has at most $\lfloor n^2/4 \rfloor$ edges. Moreover, Erdős proved that if $G$ is further non-bipartite, then $e(G)\le \lfloor {(n-1)^2}/{4}\rfloor +1$. Recently, Lin, Ning and Wu [Combin. Probab. Comput. 30 (2021)] established a spectral version by showing that if $G$ is a triangle-free non-bipartite graph on $n$ vertices, then $λ(G)\le λ(S_1(T_{n-1,2}))$, with equality if and only if $G=S_1(T_{n-1,2})$, where $S_1(T_{n-1,2})$ is obtained from $T_{n-1,2}$ by subdividing an edge. In this paper, we investigate the maximum spectral radius of a non-bipartite graph without some short odd cycles. Let $C_{2\ell +1}(T_{n-2\ell, 2})$ be the graph obtained by identifying a vertex of $C_{2\ell+1}$ and a vertex of the smaller partite set of $T_{n-2\ell ,2}$. We prove that for $1\le \ell < k$ and $n\ge 187k$, if $G$ is an $n$-vertex $\{C_3,\ldots ,C_{2\ell -1},C_{2k+1}\}$-free non-bipartite graph, then $λ(G)\le λ(C_{2\ell +1}(T_{n-2\ell, 2}))$, with equality if and only if $G=C_{2\ell +1}(T_{n-2\ell, 2})$. This result could be viewed as a spectral analogue of a min-degree result due to Yuan and Peng [European J. Combin. 127 (2025)]. Moreover, our result extends a result of Guo, Lin and Zhao [Linear Algebra Appl. 627 (2021)] as well as a recent result of Zhang and Zhao [Discrete Math. 346 (2023)] since we can get rid of the condition that $n$ is sufficiently large. The argument in our proof is quite different and makes use of the classical spectral stability method and the double-eigenvector technique. The main innovation lies in a more clever argument that guarantees a subgraph to be bipartite after removing few vertices, which may be of independent interest.