# `GenerationTheorem.lean` — Informal Summary > **Source**: [`HilleYosida/Future/GenerationTheorem.lean`](../../../HilleYosida/Future/GenerationTheorem.lean) > **Generated**: 2026-03-27 > **Note**: Auto-generated by `/lean-summarize`. Re-run to refresh. ## Overview This file states the **converse direction of the Hille-Yosida theorem** as axioms. While the forward direction (semigroup implies resolvent bound) is fully proved in `StronglyContinuousSemigroup.lean`, the converse (operator generates a semigroup) requires the Yosida approximation and is stated here for use by downstream projects. It also defines dissipative operators and the Lumer-Phillips characterization. ## Status **Main result**: Not yet proven (2 axioms) - `StronglyContinuousSemigroup.domain_isDense` (axiom, line 53) - `hilleYosidaGeneration` (axiom, line 98) **Length**: 109 lines, 2 definitions/structures + 2 axioms --- ## Key declarations ### `StronglyContinuousSemigroup.domain_isDense` (line 53) --- Axiom The generator domain $D(A)$ is dense in $X$. The proof (not yet formalized) uses "smoothed" elements $x_t = (1/t) \int_0^t S(s)x\, ds \in D(A)$ which converge to $x$ as $t \to 0$. Ref: [EN] Thm. II.1.4. ### `DenseLinearOperator` (line 65) --- Structure A densely defined linear operator $A$ on a Banach space $X$: a submodule `domain`, a linear map `op : domain -> X`, and a proof that `domain` is dense. ### `DenseLinearOperator.IsDissipative` (line 77) --- Definition An operator $A$ is dissipative if $\lVert (\lambda I - A)x \rVert \ge \lambda \lVert x \rVert$ for all $\lambda > 0$ and $x \in D(A)$. Ref: [EN] Def. II.3.13. ### `hilleYosidaGeneration` (line 98) --- Axiom **Hille-Yosida generation theorem** (contraction case, converse). A densely defined, dissipative operator $A$ with $\operatorname{rg}(\lambda_0 - A) = X$ for some $\lambda_0 > 0$ generates a contraction semigroup $S$ such that the generator of $S$ extends $A$. This is the Lumer-Phillips theorem ([EN] Thm. II.3.15). The proof (to be formalized) constructs the semigroup via the Yosida approximation $A_\lambda = \lambda^2 R(\lambda, A) - \lambda I$ and shows $e^{t A_\lambda}$ converges strongly as $\lambda \to \infty$. Ref: [EN] Thm. II.3.5, Cor. II.3.6, Thm. II.3.15; [Linares] Thm. 6. --- *This file has **2** definitions/structures and **2** axioms (both unproved).*