© 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. 22-TRANSCODER-HP
    7. 87353
    Prev
    Next
    INDEX
    Explanations

    Text excerpts

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    gst
    -0.27
    sh
    -0.25
     perfect
    -0.25
     Grund
    -0.25
    rg
    -0.25
    shr
    -0.25
    kest
    -0.25
    Round
    -0.24
     health
    -0.24
    OCK
    -0.24
    POSITIVE LOGITS
    uve
    0.26
    èĮ¸
    0.26
     Renders
    0.24
     Hiro
    0.23
    hoe
    0.23
    ár
    0.22
    &w
    0.22
    atever
    0.22
    hiro
    0.22
    hill
    0.22
    Activations Density 0.009%

    No Known Activations