Problems
Every problem posed by Paul Erdős, with its status, references, discussion and proof claims.
Every problem posed by Paul Erdős, with its status, references, discussion and proof claims.
110 of 1,221 problems match
Every problem posed by Paul Erdős, with its status, references, discussion and proof claims.
Asks whether a family of sets closed under taking subsets always has an element contained in as many members as its largest intersecting subfamily has sets.
Asks for an asymptotic formula for the least such that holds distinct with for each . It is known to lie between and , with a 2026 claim pending.
Asks whether a graph of infinite chromatic number must contain a subgraph of the same chromatic number with no odd cycle of length at most .
Asks whether a graph on vertices with diameter , in which deleting any edge raises the diameter, has at most edges. Füredi proved this for all large , and one unreviewed proof claim covers every .
Asks whether the proportion of such that every prime has a divisor of with is .
Asks for bounds on the least such that the -dimensional unit cube splits into homothetic cubes for every , and in particular whether .
Asks how large must be, as a function of , for some antichain of subsets of that uses each occurring size at least times to have distinct sizes; three claims are pending, none accepted.
Asks whether, for every , some of density above has equal products of distinct elements only with equally many factors.
Asks for the least value, over , of plus the size of a largest whose sums of distinct pairs avoid . Choi's interval function lies between and , and an unreviewed 2026 claim asserts the conjectured .
Estimates , the size of the largest subset guaranteed inside any integers in which equal sums of elements can occur only between equally many terms.
Estimates , the size of the largest subset guaranteed inside any integers in which no element equals the sum of two or more other distinct elements. Known bounds are .
Asks whether for some constant , where is the size of the largest in which every number has fewer than representations with in .
Asks whether a graph with vertices and edges has two vertices of equal degree joined by a path of length , corrected to since (the triangle) fails. Proved for by Chen and Ma; claimed for every by Liu and Zeng.
Estimates , the least for which some -element subset of has a set of subset sums with no -term progression, and asks whether . The estimate is open, and a 2026 preprint claims a negative answer to the question.
Asks whether, for every , infinitely many have more than integers with . A proof is claimed by the OpenAI release of September 2026.
Asks whether an intersecting -uniform hypergraph with chromatic number must have vertices, and whether two of its edges must meet in vertices. Plasma AI's internal whole-statement reviews with separate passing grades support the negative vertex-count answer and the positive linear-intersection answer under the exact-chromatic formulation; both claims await the sign-off of Jacob Parish at Plasma AI.
Asks whether the maximum size of a set with never squarefree for is attained by the integers .
Asks whether, for , representation counts with a suitable constant force an additive basis of order to contain a minimal one.
Asks how small the gaps of an infinite sum-free set of naturals can be, and whether the th gap can stay below .
Asks whether a set above the triangle threshold of the coprime graph forces all odd cycles up to length and, for large , complete tripartite subgraphs. Sárközy settled the second in 1999.
Asks whether there is a transcendental entire function such that, for every infinite sequence , the zeros of the derivatives together are dense in the plane.
Asks for the largest number of unit distances among disjoint translates of a compact convex set, in particular whether for some .
Asks whether some makes the least prime exceed for residues , for all large . Two 2026 proof claims answer yes, one with a Lean development; both are pending and unreviewed.
Asks whether there is such that, for every , some complex with satisfy for all .
Asks for estimates of the least such that every -element subset of has more than members divisible only by primes from some set of primes, and whether .