diff --git a/README.md b/README.md
index aea4b762f6546b31bcd583296e28408629186ec9..a4c3f1ace376cfb45d60a9a1b0ccc17caa6c7f51 100644
--- a/README.md
+++ b/README.md
@@ -1,7 +1,7 @@
-# Setup galaxy simulations from Ploeckinger et al (2023, subm.)
+# Setup galaxy simulations from Ploeckinger et al ([arXiv](https://arxiv.org/abs/2310.10721))
 
 ## Description
-Reproduce the isolated disk galaxy simulations from Ploeckinger et al. (subm. 2023).
+Reproduce the isolated disk galaxy simulations from Ploeckinger et al. ([arXiv](https://arxiv.org/abs/2310.10721)).
 
 ![Isolated disk galaxy](isolated_galaxy_examples.png "Stellar mass surface densities")
 
@@ -9,9 +9,9 @@ Reproduce the isolated disk galaxy simulations from Ploeckinger et al. (subm. 20
 ## Preparation
 
 ### Step 1: Install SWIFT
-The isolated galaxy simulations presented in Ploeckinger et al. (subm. 2023) use the [SWIFT](www.swiftsim.com) code (Schaller et al. [arXiv](https://arxiv.org/abs/2305.13380)). For reproducing these simulations install SWIFT following the information [here](https://swift.strw.leidenuniv.nl/docs/GettingStarted/index.html).
+The isolated galaxy simulations presented in Ploeckinger et al. ([arXiv](https://arxiv.org/abs/2310.10721)) use the [SWIFT](www.swiftsim.com) code (Schaller et al. [arXiv](https://arxiv.org/abs/2305.13380)). For reproducing these simulations install SWIFT following the information [here](https://swift.strw.leidenuniv.nl/docs/GettingStarted/index.html).
 
-This repository contains the parameter file used in these simulations. Note, that in the future, the required parameters for the individual modules may change. After installing all [dependencies](https://swift.strw.leidenuniv.nl/docs/GettingStarted/compiling_code.html#dependencies) of SWIFT successfully, the SWIFT version used in Ploeckinger et al. (subm. 2023) can be setup and compiled with
+This repository contains the parameter file used in these simulations. Note, that in the future, the required parameters for the individual modules may change. After installing all [dependencies](https://swift.strw.leidenuniv.nl/docs/GettingStarted/compiling_code.html#dependencies) of SWIFT successfully, the SWIFT version used in Ploeckinger et al. ([arXiv](https://arxiv.org/abs/2310.10721)) can be setup and compiled with
 
 ```
 git clone https://github.com/SWIFTSIM/SWIFT.git
@@ -42,7 +42,7 @@ In the folder `data`, run the script `getall.sh` to load all necessary data for
 
 - `./getall.sh`
 
-The `setup.py` script specifies which parameters are used. For Ploeckinger et al. (2023, subm.), we varied:
+The `setup.py` script specifies which parameters are used. For Ploeckinger et al. ([arXiv](https://arxiv.org/abs/2310.10721)), we varied:
 
 - `mass_resolution_levels`: the mass resolution (level `M5` for a particle mass of 10<sup>5</sup>M<sub>sun</sub>, and `M6` for a particle mass of 8x10<sup>5</sup>M<sub>sun</sub>)
 - `softening_resolutions_in_kpc`: the constant gravitational softening length (Plummer-equivalent softening length, &epsilon;) 
@@ -59,7 +59,7 @@ python3 setup.py
 
 ### Step 3: Rerun the simulations with very small hmin 
 
-We show in Ploeckinger et al. (2023, subm) that the SPH estimated densities are inaccurate representations of the particle positions, if the smoothing length, `h`, is limited by a minimum value, `hmin`. One of the simulation snapshots from the original simulation is converted into an initial conditions file (`data/create_ics_from_snapshot.py`) and the simulation is restarted with a very small value for `hmin` for one very short timestep. This is all done in the script `setup_reruns.py`:
+We show in Ploeckinger et al. ([arXiv](https://arxiv.org/abs/2310.10721)) that the SPH estimated densities are inaccurate representations of the particle positions, if the smoothing length, `h`, is limited by a minimum value, `hmin`. One of the simulation snapshots from the original simulation is converted into an initial conditions file (`data/create_ics_from_snapshot.py`) and the simulation is restarted with a very small value for `hmin` for one very short timestep. This is all done in the script `setup_reruns.py`:
 
 - Select the mass resolution level(s) that should be re-run and the snapshot number (default: `snapshotnumber = 100`, t = 1 Gyr).
 
@@ -87,7 +87,7 @@ If all simulation have finished successfully, the following folders and files sh
 - `M6_reruns_snap0100`: a subfolder is created for each run in `M6` with the prefix `Rerun` (e.g. `M6_reruns_snap0100/RerunGalaxyM6_soft0250pc_hmin007.75pc_sfe00.003`) containing the output file produced after rerunning the simulation for one short timestep (`output_0000.hdf5`)
 - `M6_fofruns_snap0100`: a subfolder is created for each run in `M6` with the prefix `FOF` (e.g. `M6_reruns_snap0100/FOFGalaxyM6_soft0250pc_hmin007.75pc_sfe00.003`) containing the output file `fof_output_0000.hdf5`
 
-**If all these files are present, all figures from Ploeckinger et al. (2023, subm) can be reproduced with the plotting scripts from here.**
+**If all these files are present, all figures from Ploeckinger et al. ([arXiv](https://arxiv.org/abs/2310.10721)) can be reproduced with the plotting scripts from here.**
 
 ## Authors and acknowledgment
 S. Ploeckinger, University of Vienna