diff --git a/2025/somd25.html b/2025/somd25.html
index b8a2d97..19ec82c 100644
--- a/2025/somd25.html
+++ b/2025/somd25.html
@@ -76,7 +76,7 @@ <h2>Dataset</h2>
   <p>We will upload the dataset shortly after the competition begins. </p>
 
   <h2>Evaluation</h2>
-  <p>We evaluate submissions using the F1 score, a metric that reflects the accuracy and precision of the Named Entity Recognition (NER) and Relation Extraction (RE). We will calculate an <em>exact match macro average F1 score</em> for each of the two test phases.  </p>
+  <p>We evaluate submissions using the F1 score, a metric that reflects the accuracy and precision of the Named Entity Recognition (NER) and Relation Extraction (RE). We will calculate <a href='#ref8'>macro-average F1 score using exact match</a> criteria for each of the two test phases. </p>
    
   <h2>Competition Timeline Overview</h2>
   <ul>
@@ -129,6 +129,11 @@ <h3>References</h3>
             <cite>Zhang, Y., Zhong, V., Chen, D., Angeli, G., & Manning, C. D. (2017).</cite> Position-aware Attention and Supervised Data Improve Slot Filling. 35-45. 
             <a href="https://doi.org/10.18653/v1/D17-1004">DOI:10.18653/v1/D17-1004</a>
         </li>
+      <li id="ref8">
+           <cite>Nakayama, H. (2018).</cite> seqeval: A Python framework for sequence labeling evaluation.
+<a href="https://github.com/chakki-works/seqeval">Software available from https://github.com/chakki-works/seqeval</a>
+
+        </li>
     </ul>