© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 12-TRANSCODER-HP
    7. 22328
    Prev
    Next
    INDEX
    Explanations

    Foot and stance

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_12.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.12.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    gap
    -0.33
     gap
    -0.31
    Gap
    -0.30
     gaps
    -0.28
    _gap
    -0.28
    éļĶ
    -0.26
     Gap
    -0.26
    éĹ´éļĻ
    -0.25
    -gap
    -0.25
     Flake
    -0.25
    POSITIVE LOGITS
    rego
    0.27
     smoke
    0.26
     satisfaction
    0.26
    满足
    0.26
    åĸ·
    0.25
     Smoke
    0.24
    IRO
    0.24
    uther
    0.24
    anyak
    0.24
     suce
    0.24
    Activations Density 0.001%

    No Known Activations