1: \begin{abstract}

2:    Introduced in the 1990s in the context of the algebraic approach to graph rewriting,

3:    gs-monoidal categories are symmetric monoidal categories

4:    where each object is equipped with the structure of a commutative comonoid. They arise for example as

5:    Kleisli categories of commutative monads on cartesian categories,

6:    and as such they provide a general framework for effectful computation.

7:    Recently proposed in the context of categorical probability, Markov categories are

8:    gs-monoidal categories where the monoidal unit is also terminal, and they arise for example as

9:    Kleisli categories of commutative \emph{affine} monads, where affine means that the monad preserves the monoidal unit.

10: 

11:    The aim of this paper is to study a new condition on the gs-monoidal structure, resulting in the concept of \emph{weakly Markov categories},

12:    which is intermediate between gs-monoidal categories and Markov ones.

13:    In a weakly Markov category, the morphisms to the monoidal unit are not necessarily unique, but form a group.

14:    As we show, these categories exhibit a rich theory of conditional independence for morphisms, generalising the known theory for Markov categories.

15:    We also introduce the corresponding notion for commutative monads, which we call weakly affine, and for which we give two equivalent

16:    characterisations.

17: 

18:    The paper argues that these monads are relevant to the study of categorical probability.

19:    A case at hand is the monad of finite non-zero measures, which is weakly affine but not affine.

20:    Such structures allow to investigate probability without normalisation within an elegant categorical framework.

21: \end{abstract}

22: