progressor font edit

src/pages/index.mdx

@@ -170,7 +170,7 @@ We randomly sample frame triplets triplet (<LaTeX inline formula="(\mathcal{o}_{
 
 #### Real-World Few-Shot Offline Reinforcement Learning with Noisy Demonstrations
 
-We compare PROGRESSOR with R3M and VIP by freezing the pre-trained models and using them as reward prediction models to train RWR-ACT on downstream robotic learning tasks. 
+We compare P<span style="font-size: 80%">ROGRESSOR</span> with R3M and VIP by freezing the pre-trained models and using them as reward prediction models to train RWR-ACT on downstream robotic learning tasks. 
 
 <Video source={rwr} />
 <br/>