Update methodology description to remove tuning reference

Update website/templates/benchmarks.html to remove "(tuned)" from the Cleora algorithm description in the methodology section.

Replit-Commit-Author: Agent
Replit-Commit-Session-Id: ec794acd-c4a5-47f6-b906-d70ac3c316ee
Replit-Commit-Checkpoint-Type: full_checkpoint
Replit-Commit-Event-Id: 6c3cc02c-5af6-4012-89ed-00608ee4d65a
Replit-Commit-Screenshot-Url: https://storage.googleapis.com/screenshot-production-us-central1/28ec11df-9ccf-40bc-9ff4-d0523e5b6a98/ec794acd-c4a5-47f6-b906-d70ac3c316ee/9H7rCbD
Replit-Helium-Checkpoint-Created: true

Author: JacekDabrowski1

website/templates/benchmarks.html

@@ -436,7 +436,7 @@ <h2 id="when-to-use">When to Use What</h2>
         <h2 id="methodology">Methodology</h2>
         <ul>
             <li><strong>Datasets:</strong> All datasets are downloaded from their canonical academic sources at runtime. ego-Facebook and roadNet-CA from <a href="https://snap.stanford.edu/" target="_blank">SNAP</a>. Cora, CiteSeer, and PubMed from <a href="https://github.com/kimiyoung/planetoid" target="_blank">Planetoid</a> (Yang et al., ICML 2016). PPI from <a href="https://data.dgl.ai/dataset/ppi.zip" target="_blank">DGL</a> (Zitnik &amp; Leskovec, Bioinformatics 2017). No synthetic data is used.</li>
-            <li><strong>Algorithms:</strong> Cleora, ProNE, RandNE, HOPE, NetMF, GraRep, DeepWalk, and Node2Vec. All use 256-dimensional embeddings. Cleora uses 40 iterations (tuned); all other algorithms use default hyperparameters.</li>
+            <li><strong>Algorithms:</strong> Cleora, ProNE, RandNE, HOPE, NetMF, GraRep, DeepWalk, and Node2Vec. All use 256-dimensional embeddings. Cleora uses 40 iterations; all other algorithms use default hyperparameters.</li>
             <li><strong>Labels:</strong> ego-Facebook uses Louvain community detection (seed=42). Cora/CiteSeer/PubMed use paper category labels. PPI uses functional class labels.</li>
             <li><strong>Evaluation:</strong> Nearest Centroid classifier with 80/20 train/test split (seed=42).</li>
             <li><strong>Failure modes:</strong> "OOM" means the process was killed by the OS due to exceeding available memory. "Timed Out" means the algorithm exceeded a 90-second time budget. Both indicate that the algorithm is not feasible for that dataset size.</li>