1: \begin{abstract}

2: We propose a novel method for the inference of phylogenetic trees that utilises

3: point configurations on hyperbolic space as its optimisation landscape.  Each

4: taxon corresponds to a point of the point configuration, while the evolutionary

5: distance between taxa is represented by the geodesic distance between their

6: corresponding points.  The point configuration is iteratively modified to

7: increase an objective function that additively combines pairwise log-likelihood

8: terms.  After convergence, the final tree is derived from the inter-point

9: distances using a standard distance-based method.  The objective function,

10: which is shown to mimic the log-likelihood on tree space, is a differentiable

11: function on a Riemannian manifold. Thus gradient-based optimisation techniques

12: can be applied, avoiding the need for combinatorial rearrangements of tree

13: topology.

14: \end{abstract}

15: