added report 002

_config.yml

@@ -43,7 +43,7 @@ author:
   bitbucket        : # Username - Update with your username on the site
   codepen          : # Username
   dribbble         : # Username
-  github           : "ModelOriented"
+  github           : "ModelOriented/CVE-AI"
   kaggle           : # Username
   stackoverflow    : # User number or user number and name (i.e., use "1" or "1/jeff-atwood")
 

_pages/publications.md

@@ -1,6 +1,6 @@
 ---
 layout: archive
-title: "Publications"
+title: "Reports"
 permalink: /publications/
 author_profile: true
 ---

_publications/2024-01-10-MIDI.md

@@ -1,5 +1,5 @@
 ---
-title: "Validation report - MIDI-to-score Conversion Model"
+title: "Validation report 001: MIDI-to-score Conversion Model"
 collection: publications
 permalink: /publication/2024-01-10-MIDI
 excerpt: 'This paper is about fixing template issue #693.'
@@ -25,5 +25,5 @@ more tailored and adaptable XAI methods for musical applications could contribut
 
 <hr/>
 
-Link to original publiacation with a model: https://www.turing.ac.uk/sites/default/files/2022-09/midi_quantisation_paper_ismir_2022_0.pdf
+Link to original publication with a model: [midi_quantisation_paper_ismir_2022_0.pdf](https://www.turing.ac.uk/sites/default/files/2022-09/midi_quantisation_paper_ismir_2022_0.pdf)
 

_publications/2024-01-11-GO.md

@@ -0,0 +1,26 @@
+---
+title: "Validation report 002: Go Policy Networks"
+collection: publications
+permalink: /publication/2024-01-11-GO
+excerpt: 'This paper is about fixing template issue #693.'
+date: 2024-01-11
+venue: 'Explainable Machine Learning 2023/2024 course'
+paperurl: 'https://modeloriented.github.io/CVE-AI/files/2023_GO.pdf'
+citation: 'Antoni Hanke, Michal Grotkowski. (2024). &quot;Comparative Analysis of Convolutional and Transformer Architectures in Go Policy Networks.&quot; <i>Github: CVE-AI/ModelOriented</i>.'
+tags:
+  - Go Policy Network
+  - Convolution
+  - Transformer
+---
+
+In this report we aim to spot shortcomings of using a convolutional architecture as a Go policy network. By comparing it to an equivalently trained Transformer policy and employing XAI methods such as Ceteris Paribus, we can see where each network under and overperforms. This work points in the direction of further research of Transformer architectures in positional games such as Go, where previously it was believed Convolutions were SoTA.
+
+<hr/>
+
+Our experiments show the limitations of convolutional policy networks. They are great at capturing local features, however limited in attending to global phenomena. In some tasks, like the game of Go, this might be detrimental. Transformers on the other hand are able to seamlessly incorporate both the local and global understanding via its flexible attention mechanism. This points in the direction of future work on implementing Transformers as policy networks for large positional games.
+
+
+<hr/>
+
+Link to original publication with a model: (see the report)
+