proofs
Provedlower-boundadaptiveexpected-workheadline-theorem

Expected-probe e^R threshold: under the plant hypotheses, every adaptive exactly unbiased sampler with a history-determined stopping time has E[τ₀] ≥ min(1/2, 1/(64V))·(1+e^R)/2

For all n,k∈N\htmlData{mx=1}{n, k \in \htmlData{mx=0}{\mathbb{N}}}, Ω∈Type⁡\htmlData{mx=1}{\Omega \in \htmlData{mx=0}{\operatorname{Type}}}, w:Ω→R\htmlData{mx=3}{w : \htmlData{mx=0}{\Omega} \mathrel{\htmlData{mx=2}{\to}} \htmlData{mx=1}{\mathbb{R}}}, g:N→Ω→List⁡ R→Fin⁡ n\htmlData{mx=11}{g : \htmlData{mx=0}{\mathbb{N}} \mathrel{\htmlData{mx=10}{\to}} \htmlData{mx=1}{\Omega} \mathrel{\htmlData{mx=9}{\to}} \htmlData{mx=4}{\htmlData{mx=2}{\operatorname{List}}\,\htmlData{mx=3}{\mathbb{R}}} \mathrel{\htmlData{mx=8}{\to}} \htmlData{mx=7}{\htmlData{mx=5}{\operatorname{Fin}}\,\htmlData{mx=6}{n}}}, out:Ω→List⁡ R→R\htmlData{mx=7}{\mathit{out} : \htmlData{mx=0}{\Omega} \mathrel{\htmlData{mx=6}{\to}} \htmlData{mx=3}{\htmlData{mx=1}{\operatorname{List}}\,\htmlData{mx=2}{\mathbb{R}}} \mathrel{\htmlData{mx=5}{\to}} \htmlData{mx=4}{\mathbb{R}}}, hist:(Fin⁡ n→R)→Ω→N→List⁡ R\htmlData{mx=13}{\mathit{hist} : (\htmlData{mx=2}{\htmlData{mx=0}{\operatorname{Fin}}\,\htmlData{mx=1}{n}} \mathrel{\htmlData{mx=4}{\to}} \htmlData{mx=3}{\mathbb{R}}) \mathrel{\htmlData{mx=12}{\to}} \htmlData{mx=5}{\Omega} \mathrel{\htmlData{mx=11}{\to}} \htmlData{mx=6}{\mathbb{N}} \mathrel{\htmlData{mx=10}{\to}} \htmlData{mx=9}{\htmlData{mx=7}{\operatorname{List}}\,\htmlData{mx=8}{\mathbb{R}}}}, τ:(Fin⁡ n→R)→Ω→N\htmlData{mx=9}{\tau : (\htmlData{mx=2}{\htmlData{mx=0}{\operatorname{Fin}}\,\htmlData{mx=1}{n}} \mathrel{\htmlData{mx=4}{\to}} \htmlData{mx=3}{\mathbb{R}}) \mathrel{\htmlData{mx=8}{\to}} \htmlData{mx=5}{\Omega} \mathrel{\htmlData{mx=7}{\to}} \htmlData{mx=6}{\mathbb{N}}}, E∈R\htmlData{mx=1}{E \in \htmlData{mx=0}{\mathbb{R}}} and R,V∈R\htmlData{mx=1}{R, V \in \htmlData{mx=0}{\mathbb{R}}}, if ∀ω, 0≤w(ω)\htmlData{mx=4}{\forall \omega,\ 0 \mathrel{\htmlData{mx=3}{\leq}} \htmlData{mx=2}{\htmlData{mx=0}{w}(\htmlData{mx=1}{\omega})}}, ∑ωw(ω)=1\htmlData{mx=3}{\sum_{\omega} \htmlData{mx=2}{\htmlData{mx=0}{w}(\htmlData{mx=1}{\omega})}} \mathrel{\htmlData{mx=4}{=}} 1, ∀y,ω, hist(y,ω,0)=[]\htmlData{mx=6}{\forall y, \omega,\ \htmlData{mx=3}{\htmlData{mx=0}{\mathit{hist}}(\htmlData{mx=1}{y}, \htmlData{mx=2}{\omega}, 0)} \mathrel{\htmlData{mx=5}{=}} \htmlData{mx=4}{\left[ \right]}}, ∀y,ω,t, hist(y,ω,t+1)=hist(y,ω,t)++([y(g(t,ω,hist(y,ω,t)))])\htmlData{mx=26}{\forall y, \omega, t,\ \htmlData{mx=5}{\htmlData{mx=0}{\mathit{hist}}(\htmlData{mx=1}{y}, \htmlData{mx=2}{\omega}, \htmlData{mx=4}{\htmlData{mx=3}{t} + 1})} \mathrel{\htmlData{mx=25}{=}} \htmlData{mx=24}{\htmlData{mx=10}{\htmlData{mx=6}{\mathit{hist}}(\htmlData{mx=7}{y}, \htmlData{mx=8}{\omega}, \htmlData{mx=9}{t})} + \htmlData{mx=23}{\mathtt{+}(\htmlData{mx=22}{\left[ \htmlData{mx=21}{\htmlData{mx=11}{y}(\htmlData{mx=20}{\htmlData{mx=12}{g}(\htmlData{mx=13}{t}, \htmlData{mx=14}{\omega}, \htmlData{mx=19}{\htmlData{mx=15}{\mathit{hist}}(\htmlData{mx=16}{y}, \htmlData{mx=17}{\omega}, \htmlData{mx=18}{t})})})} \right]})}}}, ∀y,y′,ω, (∀t, t≤τ(y,ω)  ⟹  hist(y′,ω,t)=hist(y,ω,t))  ⟹  τ(y′,ω)=τ(y,ω)\htmlData{mx=29}{\forall y, y', \omega,\ (\htmlData{mx=18}{\forall t,\ \htmlData{mx=0}{t} \mathrel{\htmlData{mx=5}{\leq}} \htmlData{mx=4}{\htmlData{mx=1}{\tau}(\htmlData{mx=2}{y}, \htmlData{mx=3}{\omega})} \mathrel{\htmlData{mx=17}{\implies}} \htmlData{mx=10}{\htmlData{mx=6}{\mathit{hist}}(\htmlData{mx=7}{y'}, \htmlData{mx=8}{\omega}, \htmlData{mx=9}{t})} \mathrel{\htmlData{mx=16}{=}} \htmlData{mx=15}{\htmlData{mx=11}{\mathit{hist}}(\htmlData{mx=12}{y}, \htmlData{mx=13}{\omega}, \htmlData{mx=14}{t})}}) \mathrel{\htmlData{mx=28}{\implies}} \htmlData{mx=22}{\htmlData{mx=19}{\tau}(\htmlData{mx=20}{y'}, \htmlData{mx=21}{\omega})} \mathrel{\htmlData{mx=27}{=}} \htmlData{mx=26}{\htmlData{mx=23}{\tau}(\htmlData{mx=24}{y}, \htmlData{mx=25}{\omega})}}, E=∑ωw(ω)⋅τ(⋅↦0,ω)\htmlData{mx=0}{E} \mathrel{\htmlData{mx=11}{=}} \htmlData{mx=10}{\sum_{\omega} \htmlData{mx=9}{\htmlData{mx=3}{\htmlData{mx=1}{w}(\htmlData{mx=2}{\omega})} \cdot \htmlData{mx=8}{\htmlData{mx=7}{\htmlData{mx=4}{\tau}(\htmlData{mx=5}{\cdot \mapsto 0}, \htmlData{mx=6}{\omega})}}}}, ∑ωw(ω)⋅out(ω,hist(⋅↦0,ω,τ(⋅↦0,ω)))=0\htmlData{mx=15}{\sum_{\omega} \htmlData{mx=14}{\htmlData{mx=2}{\htmlData{mx=0}{w}(\htmlData{mx=1}{\omega})} \cdot \htmlData{mx=13}{\htmlData{mx=3}{\mathit{out}}(\htmlData{mx=4}{\omega}, \htmlData{mx=12}{\htmlData{mx=5}{\mathit{hist}}(\htmlData{mx=6}{\cdot \mapsto 0}, \htmlData{mx=7}{\omega}, \htmlData{mx=11}{\htmlData{mx=8}{\tau}(\htmlData{mx=9}{\cdot \mapsto 0}, \htmlData{mx=10}{\omega})})})}}} \mathrel{\htmlData{mx=16}{=}} 0, ∑ωw(ω)⋅(out(ω,hist(⋅↦0,ω,τ(⋅↦0,ω)))−0)2≤V\htmlData{mx=17}{\sum_{\omega} \htmlData{mx=16}{\htmlData{mx=2}{\htmlData{mx=0}{w}(\htmlData{mx=1}{\omega})} \cdot \htmlData{mx=15}{(\htmlData{mx=14}{\htmlData{mx=13}{\htmlData{mx=3}{\mathit{out}}(\htmlData{mx=4}{\omega}, \htmlData{mx=12}{\htmlData{mx=5}{\mathit{hist}}(\htmlData{mx=6}{\cdot \mapsto 0}, \htmlData{mx=7}{\omega}, \htmlData{mx=11}{\htmlData{mx=8}{\tau}(\htmlData{mx=9}{\cdot \mapsto 0}, \htmlData{mx=10}{\omega})})})} - 0})^{2}}}} \mathrel{\htmlData{mx=19}{\leq}} \htmlData{mx=18}{V}, ∀S∈powersetCard⁡(k+1,univ⁡), ∑ωw(ω)⋅out(ω,hist(j↦{Rif j∈S0otherwise,ω,τ(j↦{Rif j∈S0otherwise,ω)))=k+1⋅eRk+1⋅eR+(n−(k+1))\htmlData{mx=52}{\forall S \in \htmlData{mx=51}{\htmlData{mx=46}{\operatorname{powersetCard}}(\htmlData{mx=48}{\htmlData{mx=47}{k} + 1}, \htmlData{mx=50}{\htmlData{mx=49}{\operatorname{univ}}})},\ \htmlData{mx=25}{\sum_{\omega} \htmlData{mx=24}{\htmlData{mx=2}{\htmlData{mx=0}{w}(\htmlData{mx=1}{\omega})} \cdot \htmlData{mx=23}{\htmlData{mx=3}{\mathit{out}}\left(\htmlData{mx=4}{\omega}, \htmlData{mx=22}{\htmlData{mx=5}{\mathit{hist}}\left(\htmlData{mx=11}{j \mapsto \htmlData{mx=10}{\begin{cases} \htmlData{mx=6}{R} & \text{if } \htmlData{mx=7}{j} \mathrel{\htmlData{mx=9}{\in}} \htmlData{mx=8}{S} \\ 0 & \text{otherwise} \end{cases}}}, \htmlData{mx=12}{\omega}, \htmlData{mx=21}{\htmlData{mx=13}{\tau}\left(\htmlData{mx=19}{j \mapsto \htmlData{mx=18}{\begin{cases} \htmlData{mx=14}{R} & \text{if } \htmlData{mx=15}{j} \mathrel{\htmlData{mx=17}{\in}} \htmlData{mx=16}{S} \\ 0 & \text{otherwise} \end{cases}}}, \htmlData{mx=20}{\omega}\right)}\right)}\right)}}} \mathrel{\htmlData{mx=45}{=}} \htmlData{mx=44}{\frac{\htmlData{mx=31}{\htmlData{mx=28}{\htmlData{mx=27}{\htmlData{mx=26}{k} + 1}} \cdot \htmlData{mx=30}{e^{\htmlData{mx=29}{R}}}}}{\htmlData{mx=43}{\htmlData{mx=37}{\htmlData{mx=34}{\htmlData{mx=33}{\htmlData{mx=32}{k} + 1}} \cdot \htmlData{mx=36}{e^{\htmlData{mx=35}{R}}}} + (\htmlData{mx=42}{\htmlData{mx=39}{\htmlData{mx=38}{n}} - (\htmlData{mx=41}{\htmlData{mx=40}{k} + 1})})}}}}, ∀S∈powersetCard⁡(k+1,univ⁡), ∑ωw(ω)⋅(out(ω,hist(j↦{Rif j∈S0otherwise,ω,τ(j↦{Rif j∈S0otherwise,ω)))−k+1⋅eRk+1⋅eR+(n−(k+1)))2≤V\htmlData{mx=55}{\forall S \in \htmlData{mx=54}{\htmlData{mx=49}{\operatorname{powersetCard}}(\htmlData{mx=51}{\htmlData{mx=50}{k} + 1}, \htmlData{mx=53}{\htmlData{mx=52}{\operatorname{univ}}})},\ \htmlData{mx=46}{\sum_{\omega} \htmlData{mx=45}{\htmlData{mx=2}{\htmlData{mx=0}{w}(\htmlData{mx=1}{\omega})} \cdot \htmlData{mx=44}{\left(\htmlData{mx=43}{\htmlData{mx=23}{\htmlData{mx=3}{\mathit{out}}\left(\htmlData{mx=4}{\omega}, \htmlData{mx=22}{\htmlData{mx=5}{\mathit{hist}}\left(\htmlData{mx=11}{j \mapsto \htmlData{mx=10}{\begin{cases} \htmlData{mx=6}{R} & \text{if } \htmlData{mx=7}{j} \mathrel{\htmlData{mx=9}{\in}} \htmlData{mx=8}{S} \\ 0 & \text{otherwise} \end{cases}}}, \htmlData{mx=12}{\omega}, \htmlData{mx=21}{\htmlData{mx=13}{\tau}\left(\htmlData{mx=19}{j \mapsto \htmlData{mx=18}{\begin{cases} \htmlData{mx=14}{R} & \text{if } \htmlData{mx=15}{j} \mathrel{\htmlData{mx=17}{\in}} \htmlData{mx=16}{S} \\ 0 & \text{otherwise} \end{cases}}}, \htmlData{mx=20}{\omega}\right)}\right)}\right)} - \htmlData{mx=42}{\frac{\htmlData{mx=29}{\htmlData{mx=26}{\htmlData{mx=25}{\htmlData{mx=24}{k} + 1}} \cdot \htmlData{mx=28}{e^{\htmlData{mx=27}{R}}}}}{\htmlData{mx=41}{\htmlData{mx=35}{\htmlData{mx=32}{\htmlData{mx=31}{\htmlData{mx=30}{k} + 1}} \cdot \htmlData{mx=34}{e^{\htmlData{mx=33}{R}}}} + (\htmlData{mx=40}{\htmlData{mx=37}{\htmlData{mx=36}{n}} - (\htmlData{mx=39}{\htmlData{mx=38}{k} + 1})})}}}}\right)^{2}}}} \mathrel{\htmlData{mx=48}{\leq}} \htmlData{mx=47}{V}}, k+1≤n\htmlData{mx=1}{\htmlData{mx=0}{k} + 1} \mathrel{\htmlData{mx=3}{\leq}} \htmlData{mx=2}{n}, 0<V0 \mathrel{\htmlData{mx=1}{<}} \htmlData{mx=0}{V}, 0≤R0 \mathrel{\htmlData{mx=1}{\leq}} \htmlData{mx=0}{R}, n−(k+1)≤k+1⋅eR\htmlData{mx=4}{\htmlData{mx=1}{\htmlData{mx=0}{n}} - (\htmlData{mx=3}{\htmlData{mx=2}{k} + 1})} \mathrel{\htmlData{mx=11}{\leq}} \htmlData{mx=10}{\htmlData{mx=7}{\htmlData{mx=6}{\htmlData{mx=5}{k} + 1}} \cdot \htmlData{mx=9}{e^{\htmlData{mx=8}{R}}}} and k+1⋅(1+eR)≤2n\htmlData{mx=6}{\htmlData{mx=2}{\htmlData{mx=1}{\htmlData{mx=0}{k} + 1}} \cdot (\htmlData{mx=5}{1 + \htmlData{mx=4}{e^{\htmlData{mx=3}{R}}}})} \mathrel{\htmlData{mx=9}{\leq}} \htmlData{mx=8}{2\htmlData{mx=7}{n}} (given Fintype⁡ Ω\htmlData{mx=2}{\htmlData{mx=0}{\operatorname{Fintype}}\,\htmlData{mx=1}{\Omega}}), then

min⁡(12,164V)⋅(1+eR2)≤E\htmlData{mx=9}{\htmlData{mx=4}{\min(\htmlData{mx=0}{\frac{1}{2}}, \htmlData{mx=3}{\frac{1}{\htmlData{mx=2}{64\htmlData{mx=1}{V}}}})} \cdot \left(\htmlData{mx=8}{\frac{\htmlData{mx=7}{1 + \htmlData{mx=6}{e^{\htmlData{mx=5}{R}}}}}{2}}\right)} \mathrel{\htmlData{mx=11}{\leq}} \htmlData{mx=10}{E}
What the notation means · or point at any symbol
NNatural numbers
The natural numbers 0, 1, 2, …. In Lean they start at 0. (ℕ)
x ∈ TVariables
n, k range over natural numbers: the statement must hold for every choice. ((n k : ℕ))
Type
Type: a definition from the library, not special notation. Its name is its meaning; open it in Lean to see how it is defined. (Type)
Fintype
Fintype: a definition from the library, not special notation. Its name is its meaning; open it in Lean to see how it is defined. (Fintype)
f(x)Function application
Fintype applied to Ω. Lean writes f x; papers write f(x). (Fintype Ω)
⟹Implies
If Ω, then ℝ. (Between types, A → B is the type of functions from A to B.) (Ω → ℝ)
∀For all
The statement after the comma holds for every ω. (∀ ω, 0 ≤ w ω)
∑Summation
Add up w(ω) for every ω. (∑ ω, w ω)
=Equality
∑ ω, w(ω) and 1 are the same. (∑ ω, w ω = 1)
<Less than
0 is strictly smaller than V. (0 < V)
minMinimum
The smaller of 1/2 and 1/(64 · V). (min (1 / 2) (1 / (64 * V)))
RReal numbers
The real numbers: every point on the number line. (ℝ)
≤At most
0 is less than or equal to w(ω). (0 ≤ w ω)
powersetCard
Finset.powersetCard: a definition from the library, not special notation. Its name is its meaning; open it in Lean to see how it is defined. (Finset.powersetCard)
List
List: a definition from the library, not special notation. Its name is its meaning; open it in Lean to see how it is defined. (List)
·Multiplication
w(ω) times τ(_ ↦ 0, ω). (w ω * (τ (fun _ => 0) ω : ℝ))
aᵇPower
out(ω, hist(_ ↦ 0, ω, τ(_ ↦ 0, ω))) − 0 raised to the power 2. ((out ω (hist (fun _ => 0) ω (τ (fun _ =…)
-Subtraction
out(ω, hist(_ ↦ 0, ω, τ(_ ↦ 0, ω))) minus 0. On ℕ, Lean's subtraction stops at 0: 2 − 5 = 0. (out ω (hist (fun _ => 0) ω (τ (fun _ =>…)
(x : T)As real numbers
τ(_ ↦ 0, ω) read as an element of real numbers. Mathematicians leave this implicit; Lean needs it to know which kind of number is meant. ((τ (fun _ => 0) ω : ℝ))
Fin
Fin: a definition from the library, not special notation. Its name is its meaning; open it in Lean to see how it is defined. (Fin)
+Addition
t plus 1. (t + 1)
x ↦ f(x)Function (λ-abstraction)
The function sending _ to 0. Lean writes fun x => …; papers write x ↦ …. (fun _ => 0)
[a, b]List
The list of , in order. ([])
(x : T)Type ascription
univ read as a Finset(Fin(n)). Mathematicians leave this implicit; Lean needs it to know which kind of number is meant. ((Finset.univ : Finset (Fin n)))
univ
Finset.univ: a definition from the library, not special notation. Its name is its meaning; open it in Lean to see how it is defined. (Finset.univ)
{ a if c; b otherwiseCases (if … then … else)
R when j ∈ S, and 0 otherwise. (if j ∈ S then R else 0)
∈Element of
j is a member of S. (j ∈ S)
a/bDivision
(k + 1) · exp(R) divided by (k + 1) · exp(R) + (n − (k + 1)). In Lean, dividing by 0 gives 0. (((k + 1 : ℝ) * Real.exp R) / ((k + 1 : …)
eˣExponential function
e raised to the power R, where e ≈ 2.718. (Real.exp R)

The expected-work form of the two-sided sampling threshold, closing the fixed-budget-versus-expected-probe gap stated as open in whitepaper v3 (suggested by an external reviewer, 2026-10-05). Same instance family as ua-adaptive-sampling-threshold-explicit (unplanted row of all zeros; planted rows with score R on a (k+1)-subset), same model (coordinate probes with shared finite randomness, no preprocessing), but the probe count is a random stopping time with expectation E = Σ_ω w_ω τ(0, ω). One statement: plant-size existence, the planted output ≥ 1/2, and E ≥ min(1/2, 1/(64V))·(1+e^R)/2. So the blind cost e^R is tight for expected work as well as for fixed budgets, in this model. Preprocessing is still excluded.

How it was proved

Proved
5 October 2026
By
Ajax, with the agent “Baguette math & proofs (Claude Code, 2026-10-03)”
Checked by
Lean's kernel (Lean 4.33.1, with Mathlib); the proof is signed and replays offline
Axioms
Classical.choice, Quot.sound, propext
Proof
295 lines of Lean
Sign in to see the proofThe proof, its signed report and its history.

More proved statements