Proposition 10.3.5 — finite coefficient determination
DisprovedPutermanMDP.Sensitive.finite_coefficient_determinationLet be deterministic decision rules in a finite Markov decision process. Suppose has recurrent classes and their Laurent coefficient vectors agree through the indicated order:
The printed proposition concludes
This is the chapter's finite-determination claim. FINDINGS records a counterexample at ; the statement is preserved rather than silently adding a hypothesis.
Formalization Note The threshold uses integer subtraction and is when . There is no assumption that also has recurrent classes.
import Mathlib import Definitions.Def_PutermanMDP_Sensitive_Coefficients
namespace PutermanMDP.Sensitive
/-- Printed Proposition 10.3.5. The coefficient interval is closed at both ends:
when r = |S| it consists of the single index -1; when r = |S|-1 it is {-1, 0}.
The r=|S| boundary is disputed in FINDINGS. -/
theorem finite_coefficient_determination {S A : Type} [Fintype S] [DecidableEq S] [Nonempty S] [Fintype A]
(M : Model S A) (δ d : DecisionRule M) (r : ℕ)
(hr : recurrentCount (policyMatrix d) = r)
(heq : ∀ k ∈ Set.Icc (-1 : ℤ) ((Fintype.card S : ℤ) - (r : ℤ) - 1),
coefficient δ k = coefficient d k) :
∀ k : ℤ, -1 ≤ k → coefficient δ k = coefficient d k := by sorry
end PutermanMDP.Sensitive
Read-back
What the Lean code literally says, in plain math · claude-opus-5-5
Setting. Let and be types. is finite, has decidable equality and is nonempty. is finite and may be empty. The statement takes the following data:
- , a value of type
Model S A; - and , two values of type
DecisionRule M; - , a natural number.
Model, DecisionRule, policyMatrix, recurrentCount and coefficient come from the imported file Definitions.Def_PutermanMDP_Sensitive_Coefficients. Their definitions are not part of this declaration, so this read-back cannot expand them. Read from their types alone:
- (
policyMatrix d) is some matrix-like object built from the decision rule . - (
recurrentCount) is a natural number computed from . - (
coefficient δ k) is a value attached to the decision rule and an integer index . The type of that value is not visible here. The code does not pass tocoefficientexplicitly; any dependence on comes through the type of .
Hypotheses.
- . Only 's count is constrained. The statement makes no assumption about .
- For every integer in the closed integer interval below, the coefficients agree: . The bound is computed in , so there is no truncation at zero.
Here is the number of elements of .
Conclusion. For every integer ,
The coefficients at indices up to are assumed to agree. The conclusion then says they agree at every index . Indices appear in neither the hypotheses nor the conclusion. The comment attached to the code describes the cases and and says the case is disputed elsewhere. That comment is not part of what the statement asserts.
Degenerate cases.
- . The interval is . Agreement at the single index is then claimed to force agreement at every .
- . The interval is .
- . The interval is .
- . The upper end is at most , so the interval is empty and the second hypothesis holds automatically. The statement would then assert that for all , for any and any with . Whether can actually exceed depends on the imported definition, which is not visible here.
- One-element (). The interval is . That is when , when , and empty when .
- Empty . Whether
DecisionRule Mhas any inhabitants then depends on the imported definitions. If it has none, the statement is vacuous. - Undefined arithmetic. Nothing in the statement itself divides, subtracts in , takes a supremum or integrates. Any default values arising inside
coefficient,policyMatrixorrecurrentCountare hidden in the imported definitions.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.