AFTD Auto-Formalizing Theoretical Domains

target_population_le_capacity

theoremverified

For any natural number n, target population vector y in stdSimplex ℝ (Fin n), Gibbs vector γ in stdSimplex ℝ (Fin n) with strictly positive entries (∀ i, 0 < γ i), and subset S : Finset (Fin n), the sum ∑ i ∈ S, y i is at most population_capacity γ y S.

Statement

theorem target_population_le_capacity {n : ℕ}
    {y γ : Fin n → ℝ} (hy : y ∈ stdSimplex ℝ (Fin n)) (hγ : γ ∈ stdSimplex ℝ (Fin n))
    (hγ_pos : ∀ i, 0 < γ i) (S : Finset (Fin n)) :
    (∑ i ∈ S, y i) ≤ population_capacity γ y S
Source
arXiv:2610.35433, Eq. (6)
Verified
08 Nov 2026
Axioms
Classical.choiceQuot.soundpropext
Built on 8
  • population_capacityGiven a Gibbs vector γ, an initial state x on Fin n, and a subset S of energy levels, the population capacity is the supremum over all…
  • thermally_reachable_subset_base_polytopeFor any surjective block-assignment map proj : Fin n → Fin m and vectors γ, x : Fin n → ℝ, thermally_reachable_block_populations proj γ x…
  • block_population_capacityGiven a block-assignment map proj : Fin n → Fin m, vectors γ, x : Fin n → ℝ, and a subset S : Finset (Fin m), block_population_capacity…
  • fiber_sum_vecGiven an assignment of n microstates to m energy levels, the fiber sum vector of a microstate distribution y has k-th component equal to…
  • submodular_base_polytopeThe base polytope B(f) of a set function f on Finset (Fin m) is the set of vectors z in ℝ^m such that the sum of z over every subset S is…
  • thermally_reachable_microA microstate distribution x' on n energy levels is thermally reachable from an initial distribution x with respect to the Gibbs vector γ…
  • is_gibbs_stochasticA square transition matrix T on n microstates is Gibbs-stochastic with respect to an equilibrium probability vector γ if its entries are…
  • thermally_reachable_block_populationsGiven a block-assignment map proj from Fin n to Fin m and vectors γ, x : Fin n → ℝ, thermally_reachable_block_populations proj γ x is the…

Read back from the Lean

For any natural number $n$, vectors $y, \gamma : \text{Fin } n \to \mathbb{R}$, and subset $S \subseteq \text{Fin } n$, if $y$ belongs to the standard simplex $\Delta^{n-1}$, $\gamma$ belongs to the standard simplex $\Delta^{n-1}$, and every coordinate of $\gamma$ is strictly positive ($\gamma_i > 0$ for all $i \in \text{Fin } n$), then the sum $\sum_{i \in S} y_i$ is less than or equal to $\text{population\_capacity}(\gamma, y, S)$.

Written by a model that saw only the Lean, never the English above. If the two disagree, that is worth a challenge.

The proof, explained

To show that the target population on any subset S is at most its population capacity, the proof first establishes that the target state y is thermally reachable from itself under the identity projection. This reachability is witnessed by the identity transition matrix, which is Gibbs-stochastic and leaves y invariant. Because every thermally reachable block population belongs to the submodular base polytope of the capacity function, the subset inequality holds by definition of the base polytope.

  1. The target state y can be witnessed as thermally reachable from itself using the identity transition matrix, which is verified to be Gibbs-stochastic.

    To show y ∈ thermally_reachable_block_populations id γ y, we provide y as the intermediate microstate and the identity matrix T_{i j} = if i = j then 1 else 0 as the transition matrix. We verify the three defining properties of a Gibbs-stochastic matrix: non-negativity of entries (0 ≤ 1 and 0 ≤ 0 via split_ifs and linarith), column sums equal to 1 (∑ i, δ_{i j} = 1 via simp), and invariance of the Gibbs distribution (∑ j, δ_{i j} γ_j = γ_i via simp).

    lines 3–10

      refine ⟨y, hy, ⟨fun i j => if i = j then 1 else 0, ?_, ?_⟩, ?_⟩
      · refine ⟨fun i j => ?_, fun j => ?_, fun i => ?_⟩
        · dsimp only
          split_ifs <;> linarith
        · dsimp only
          simp
        · dsimp only
          simp
  2. The identity transition matrix preserves the population vector y.

    For each index i, the sum ∑ j, δ_{i j} y_j simplifies directly to y_i by collapsing the sum at j = i.

    lines 11–13

      · intro i
        dsimp only
        simp
  3. The fiber-sum vector of y under the identity map is y itself, completing the proof that y is thermally reachable from itself.

    For each coordinate k, the fiber sum over the preimage {i | id(i) = k} is the sum over {i | i = k}. We show this fiber set equals the singleton {k}, so the sum reduces to y_k.

    lines 14–20

      · ext k
        dsimp [fiber_sum_vec]
        have : (Finset.univ.filter (fun i : Fin n => i = k)) = {k} := by
          ext i
          simp
        rw [this]
        simp
  4. The target vector y is thermally reachable from itself.

    After noting the hypotheses on γ, we apply the helper lemma target_population_le_capacity_mem_reachable to hy (which states y ∈ stdSimplex ℝ (Fin n)) to deduce that y ∈ thermally_reachable_block_populations id γ y.

    lines 27–30

    uses target_population_le_capacity_mem_reachable

      have _ := hγ
      have _ := hγ_pos
      have hy_reach : y ∈ thermally_reachable_block_populations id γ y :=
        target_population_le_capacity_mem_reachable hy
  5. The target population on any subset S is bounded above by the population capacity of S.

    Because the identity projection is surjective (Function.surjective_id), the theorem thermally_reachable_subset_base_polytope implies that y belongs to the submodular base polytope of the block population capacity. By definition of the base polytope, the sum of y over any subset S is at most the capacity block_population_capacity id γ y S, which is definitionally population_capacity γ y S.

    lines 31–32

    uses thermally_reachable_subset_base_polytope, Function.surjective_id

      have h_sub := thermally_reachable_subset_base_polytope id Function.surjective_id γ y hy_reach
      exact h_sub.1 S
target_population_le_capacity_mem_reachable
Any probability vector y in the standard simplex is thermally reachable from itself under the identity projection with respect to any Gibbs vector γ.
thermally_reachable_subset_base_polytope
For any surjective projection, the set of thermally reachable block populations is contained in the submodular base polytope of the block population capacity function.
Function.surjective_id
The identity function on any type is surjective.

Written by a model (gemini-3.8-flash) from the Lean proof and the goal states Lean reported at each step. Lean verified the proof; this explanation is a reading aid and is not itself machine-checked.

Lean source view module on GitHub

lemma target_population_le_capacity_mem_reachable {n : ℕ} {y γ : Fin n → ℝ} (hy : y ∈ stdSimplex ℝ (Fin n)) :    y ∈ thermally_reachable_block_populations id γ y := by  refine ⟨y, hy, ⟨fun i j => if i = j then 1 else 0, ?_, ?_⟩, ?_⟩  · refine ⟨fun i j => ?_, fun j => ?_, fun i => ?_⟩    · dsimp only      split_ifs <;> linarith    · dsimp only      simp    · dsimp only      simp  · intro i    dsimp only    simp  · ext k    dsimp [fiber_sum_vec]    have : (Finset.univ.filter (fun i : Fin n => i = k)) = {k} := by      ext i      simp    rw [this]    simp/-- A target state population on any subset is at most its own population capacity. -/theorem target_population_le_capacity {n : ℕ}    {y γ : Fin n → ℝ} (hy : y ∈ stdSimplex ℝ (Fin n)) (hγ : γ ∈ stdSimplex ℝ (Fin n))    (hγ_pos : ∀ i, 0 < γ i) (S : Finset (Fin n)) :    (∑ i ∈ S, y i) ≤ population_capacity γ y S := by  have _ := hγ  have _ := hγ_pos  have hy_reach : y ∈ thermally_reachable_block_populations id γ y :=    target_population_le_capacity_mem_reachable hy  have h_sub := thermally_reachable_subset_base_polytope id Function.surjective_id γ y hy_reach  exact h_sub.1 S
See it in the graphDiscuss this resultChallenge it