🤖 BERT + ELMo Sentence Similarity Analyzer

Add sentences one by one (minimum 2) and compute pairwise similarity using BERT and ELMo embeddings.
🔧 Select BERT Model

Choose between cased/uncased and base/large variants

📖 How to use:

  1. Choose Model: Select your preferred BERT variant (uncased recommended for similarity)
  2. Add sentences: Type a sentence and click "Add Sentence"
  3. Repeat: Add at least 2 sentences (you can add more!)
  4. Compute: Click "Compute Similarity" to see the results
  5. Export: Download embeddings and similarity matrices for further analysis
  6. Interpret: Values closer to 1.0 indicate higher similarity

🔬 Models:

  • BERT Large Uncased: Best for semantic similarity (recommended) - 1024 dimensions
  • BERT Large Cased: Preserves capitalization, good for proper nouns - 1024 dimensions
  • BERT Base Uncased: Faster, smaller model - 768 dimensions
  • BERT Base Cased: Cased version of base model - 768 dimensions
  • ELMo: Contextual word representations using LSTM - 1024 dimensions