© 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. 88360
    Prev
    Next
    INDEX
    Explanations

    Abbreviations for titles/credentials

    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
    inge
    -0.29
     classical
    -0.29
    ersh
    -0.26
    enco
    -0.24
     pits
    -0.23
    chn
    -0.23
    .ValidationError
    -0.23
     classic
    -0.23
    anh
    -0.23
    izens
    -0.23
    POSITIVE LOGITS
    Grand
    0.30
     Vendor
    0.30
     Grand
    0.29
    Vendor
    0.27
    grand
    0.26
     Signal
    0.26
    èĻ½
    0.25
    ÑĩÑĮ
    0.25
    right
    0.25
    ç¢į
    0.24
    Activations Density 0.019%

    No Known Activations