Alphabeta Math
CounterexampleConstruction: AI-adaptedVerification: AI-generatedSession-authored (Fable 5 assisted)precheck passjudge pass (z-ai/glm-5.2)audited 2026-07-29
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced — the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted — a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated — a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Continuous fn0f_n \to 0 pointwise on [0,1][0,1] with 01fn=1\int_0^1 f_n = 1 for every nn

Statement refuted

False claim: if (fn)(f_n) is a sequence of Riemann integrable functions on [0,1][0,1] converging pointwise to ff, and ff is integrable, then 01fn01f\int_0^1 f_n \to \int_0^1 f.

For nNn \in \mathbb{N} write cn:=ι(n+1)1c_n := \iota(n+1) \ge 1 (The canonical natural ι(n)=n1F\iota(n) = n \cdot 1_F of a field, Canonical naturals are positive and strictly increasing) and define the tent fn:[0,1]Rf_n : [0,1] \to \mathbb{R} by

fn(x)  :=  {4cn2x0x12cn,4cn2(1cnx)12cnx1cn,01cnx1.f_n(x) \;:=\; \begin{cases} 4c_n^{2}\,x & 0 \le x \le \tfrac{1}{2c_n}, \\[3pt] 4c_n^{2}\bigl(\tfrac{1}{c_n} - x\bigr) & \tfrac{1}{2c_n} \le x \le \tfrac{1}{c_n}, \\[3pt] 0 & \tfrac{1}{c_n} \le x \le 1 . \end{cases}

Each fnf_n is continuous on [0,1][0,1], hence integrable, with

01fn  =  1for every n,\int_0^1 f_n \;=\; 1 \qquad \text{for every } n ,

while fn(x)0f_n(x) \to 0 for every x[0,1]x \in [0,1]. So the pointwise limit is the zero function, whose integral is 00, and the integrals do not converge to it.

The heights are unbounded: fnf_n attains the value 2cn2c_n at x=1/(2cn)x = 1/(2c_n), and 2cn2c_n \to \infty. That is what the example refutes and what it does not: it refutes the interchange for pointwise convergence, and it says nothing whatever about sequences that are uniformly bounded, for which no theorem is stated on this page in any direction.

Facts & Assumptions

Given: For nNn \in \mathbb{N}, cn=ι(n+1)c_n = \iota(n+1) and the function fnf_n above; a point x[0,1]x \in [0,1] and a real ε>0\varepsilon > 0.

[L1]
[L6]

For n1n \ge 1 the map xxnx \mapsto x^{n} has derivative ι(n)xn1\iota(n)x^{\,n-1}; sums and scalar multiples of differentiable functions are differentiable with the corresponding derivatives; and pqc=c(qp)\int_p^q c = c(q-p) for a constant (For a natural n1n \ge 1 the function xxnx \mapsto x^{n} is differentiable everywhere with derivative ι(n)xn1\iota(n)\,x^{\,n-1}; for n=0n = 0 it is the constant 11, with derivative 00; for a natural n1n \ge 1 the function xxnx \mapsto x^{-n} is differentiable at every x0x \ne 0 with derivative ι(n)xn1-\iota(n)\,x^{-n-1}; consequently every polynomial function is differentiable at every real, with the derivative computed term by term, Sums, scalar multiples, products and quotients: (f+g)(c)=f(c)+g(c)(f+g)'(c) = f'(c) + g'(c), (αf)(c)=αf(c)(\alpha f)'(c) = \alpha f'(c), (fg)(c)=f(c)g(c)+f(c)g(c)(fg)'(c) = f'(c)g(c) + f(c)g'(c), and (f/g)(c)=(f(c)g(c)f(c)g(c))/g(c)2(f/g)'(c) = \bigl(f'(c)g(c) - f(c)g'(c)\bigr)/g(c)^{2} when g(c)0g(c) \ne 0, If mfMm \le f \le M on [a,b][a,b] then m(ba)L(f,P)abfabfU(f,P)M(ba)m(b-a) \le L(f,P) \le \underline{\int_a^b} f \le \overline{\int_a^b} f \le U(f,P) \le M(b-a) for every partition PP; in particular every constant function is integrable, with abc=c(ba)\int_a^b c = c(b-a), The derivative f(c)=limxcf(x)f(c)xcf'(c) = \lim_{x \to c} \frac{f(x) - f(c)}{x - c} of f:ARf : A \to \mathbb{R} at a point cAc \in A that is a limit point of AA, and differentiability on a set, Integer powers ama^m).

[L7]

A sequence of reals converges to LL when for every real ε>0\varepsilon>0 there is NN with anL<ε|a_n - L| < \varepsilon for all nNn \ge N (Limits and Cauchy sequences of reals, Sequences of reals: bounded, eventually, frequently, tails, subsequences).

Counterexample

technique · direct
1.1

cn1c_n \ge 1 by [L1], so 0<1/(2cn)<1/cn10 < 1/(2c_n) < 1/c_n \le 1 and the three pieces of the definition subdivide [0,1][0,1].

givenL1
2.1

The three formulas agree at the shared endpoints: at x=1/(2cn)x = 1/(2c_n) both give 4cn2/(2cn)=2cn4c_n^{2}/(2c_n) = 2c_n, and at x=1/cnx = 1/c_n the second gives 00, which is the third. So fnf_n is a well-defined function and is continuous on [0,1][0,1] by [L2], hence integrable by [L3].

step 1.1L2L3
2.2

On [1/cn,1][1/c_n, 1] the function fnf_n is constantly 00, so 1/cn1fn=0\int_{1/c_n}^{1} f_n = 0 by [L6]; when 1/cn=11/c_n = 1 this piece is degenerate and the integral is 00 by [L4].

step 1.1L4L6
2.3

Pointwise convergence to 00. At x=0x = 0 every fn(0)=0f_n(0) = 0. For x>0x > 0, [L1] gives NN with 1/ι(N+1)<x1/\iota(N+1) < x, and for nNn \ge N one has cn=ι(n+1)ι(N+1)c_n = \iota(n+1) \ge \iota(N+1), hence 1/cn1/ι(N+1)<x1/c_n \le 1/\iota(N+1) < x, so xx lies in the third piece and fn(x)=0f_n(x) = 0.

step 1.1givenL1L8
3.1

On [0,1/(2cn)][0, 1/(2c_n)] the function H(x):=2cn2x2H(x) := 2c_n^{2}x^{2} has H(x)=4cn2x=fn(x)H' (x)= 4c_n^{2}x = f_n(x) by [L6], so 01/(2cn)fn=H(1/(2cn))H(0)=2cn2/(4cn2)=1/2\int_0^{1/(2c_n)} f_n = H(1/(2c_n)) - H(0) = 2c_n^{2}/(4c_n^{2}) = 1/2 by [L5].

step 2.1L5L6
3.2

On [1/(2cn),1/cn][1/(2c_n), 1/c_n] the function H2(x):=4cn2(x/cnx2/2)H_2(x) := 4c_n^{2}\bigl(x/c_n - x^{2}/2\bigr) has H2(x)=4cn2(1/cnx)=fn(x)H_2'(x) = 4c_n^{2}(1/c_n - x) = f_n(x) by [L6], and H2(1/cn)=4cn2(1/cn21/(2cn2))=2H_2(1/c_n) = 4c_n^{2}\bigl(1/c_n^{2} - 1/(2c_n^{2})\bigr) = 2 while H2(1/(2cn))=4cn2(1/(2cn2)1/(8cn2))=3/2H_2(1/(2c_n)) = 4c_n^{2}\bigl(1/(2c_n^{2}) - 1/(8c_n^{2})\bigr) = 3/2; so 1/(2cn)1/cnfn=23/2=1/2\int_{1/(2c_n)}^{1/c_n} f_n = 2 - 3/2 = 1/2 by [L5].

step 2.1L5L6
3.3

Hence fn(x)0=0<ε|f_n(x) - 0| = 0 < \varepsilon for all nNn \ge N, so fn(x)0f_n(x) \to 0 for every x[0,1]x \in [0,1] by [L7].

step 2.3L7
4.1

By [L4] applied twice, 01fn=1/2+1/2+0=1\int_0^1 f_n = 1/2 + 1/2 + 0 = 1 for every nn.

step 3.1step 3.2step 2.2L4
5.1

The pointwise limit is the zero function, which is integrable with integral 00 by [L6], while 01fn=1\int_0^1 f_n = 1 for every nn by step 4.1; so the integrals do not converge to the integral of the limit and the claim is false.

step 4.1step 3.3L6L7

Remarks

  • What this refutes, stated exactly. It refutes the interchange of a limit with an integral under pointwise convergence alone, even when every fnf_n is continuous and the limit function is as regular as possible. It does not refute, and does not address, any statement about uniformly convergent sequences or about uniformly bounded ones; no such statement is proved on this page, and none is contradicted here.

  • Unboundedness of the heights is essential to the construction and is stated as a feature, not hidden. sup[0,1]fn=2cn\sup_{[0,1]} f_n = 2c_n grows without bound, and the mass 11 escapes into a spike of shrinking width. A reader who wants a theorem in this direction should note that the hypothesis to look for is a bound on the whole sequence, and that whichever theorem supplies it is not on this page.

  • The integral of the limit exists here. The failure is not that the limit function is non-integrable — it is the zero function — but that the numbers 01fn\int_0^1 f_n simply do not converge to 010\int_0^1 0. A separate failure, in which the pointwise limit of integrable functions is not integrable at all, is recorded as a false statement on the companion page of The Riemann Integral.

Depends on

Used by

Nothing in the library uses this result yet.

Dependency tree · next 3 levels

Direct dependencies and their dependencies through the next three levels: 130 results over 33 levels. An arrow runs from a result to what uses it, and this result sits at the bottom with a heavier outline. Click the chart to enlarge it.

Sources