© 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. 18-TRANSCODER-HP
    7. 89323
    Prev
    Next
    INDEX
    Explanations

    nor

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     refuse
    -0.26
    æĭĴ
    -0.25
    alk
    -0.25
    åģ·
    -0.24
    鸳
    -0.24
    vide
    -0.23
    _SK
    -0.23
    sil
    -0.23
     Aad
    -0.23
     timely
    -0.23
    POSITIVE LOGITS
    éĽ¢ãĤĮ
    0.30
    á¼´
    0.27
    (#)
    0.26
    /of
    0.25
    òng
    0.24
    enberg
    0.24
    ardware
    0.24
    etime
    0.23
    OpenHelper
    0.23
    otive
    0.23
    Activations Density 3.323%

    No Known Activations