/- Copyright 2026 The Formal Conjectures Authors. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at https://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. -/ import FormalConjecturesUtil

Erdős Problem 323

Reference: erdosproblems.com/323

open Filteropen scoped Asymptotics namespace Erdos323

Let $1\leq m\leq k$ and $f_{k,m}(x)$ denote the number of integers $\leq x$ which are the sum of $m$ many nonnegative $k$th powers.

noncomputable def f (k m x : ) : := { n : | n x (v : Fin m ), n = i, v i ^ k }.ncard

Is it true that $f_{k,k}(x) \gg_\epsilon x^{1-\epsilon}$ for all $\epsilon>0$?

This would have significant applications to Waring's problem. Erdős and Graham describe this as 'unattackable by the methods at our disposal'.

@[category research open, AMS 11] theorem declaration uses 'sorry'erdos_323.parts.i : answer(sorry) k 1, ε > (0 : ), (fun (x : ) (x : ) ^ (1 - ε)) =O[atTop] (fun (x : ) (f k k x : )) := True k 1, ε > 0, (fun x => x ^ (1 - ε)) =O[atTop] fun x => (f k k x) All goals completed! 🐙

Is it true that if $m < k$ then $f_{k,m}(x) \gg x^{m/k}$ for sufficiently large $x$?

@[category research open, AMS 11] theorem declaration uses 'sorry'erdos_323.parts.ii : answer(sorry) k m : , 1 m m < k (fun (x : ) (x : ) ^ ((m : ) / (k : ))) =O[atTop] (fun (x : ) (f k m x : )) := True (k m : ), 1 m m < k (fun x => x ^ (m / k)) =O[atTop] fun x => (f k m x) All goals completed! 🐙

The case $k=2$ was resolved by Landau, who showed $f_{2,2}(x) \sim \frac{cx}{\sqrt{\log x}}$ for some constant $c>0$.

@[category research solved, AMS 11] theorem declaration uses 'sorry'erdos_323.variants.k_eq_2 : c > 0, (fun (x : ) (f 2 2 x : )) ~[atTop] (fun (x : ) c * (x : ) / Real.sqrt (Real.log (x : ))) := c > 0, (fun x => (f 2 2 x)) ~[atTop] fun x => c * x / (Real.log x) All goals completed! 🐙

For $k>2$ it is not known if $f_{k,k}(x)=o(x)$.

@[category research open, AMS 11] theorem declaration uses 'sorry'erdos_323.variants.k_gt_2 : answer(sorry) k > 2, (fun (x : ) (f k k x : )) =o[atTop] (fun (x : ) (x : )) := True k > 2, (fun x => (f k k x)) =o[atTop] fun x => x All goals completed! 🐙 end Erdos323