LLM Interpretability & Superposition Probe Lab

6 Layers · 4 Heads/L · d_model=256

Transformer Residual Stream & Circuit DAG

18 Circuit Nodes Active
Prompt:
Residual Stream Attn Head MLP Layer Ablated

Sparse Autoencoder (SAE) Feature Decomposition & Steering

8x Dictionary Expansion
Extracted Monosemantic Features (Layer 4) Clamp to steer next-token output
Base Logit Winner
Paris (94.2%)
Logit Rank: #1
Entropy: 0.24 nats
Polysemantic Overlap 0.68
SAE Recon Loss 0.042
Active Sparsity (L0) 5 / 2048
Circuit Nodes 18
Interpretability Horizon: Superposition density exceeds single-neuron interpretability; SAE feature decomposition required.
Enjoy this tool? Build your own with Super