© 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. 89487
    Prev
    Next
    INDEX
    Explanations

    code

    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
     Twice
    -0.27
    å¹³åĿĩæ°´å¹³
    -0.27
    妹åŃIJ
    -0.26
    ÙĪØŃ
    -0.26
    die
    -0.25
     {}:
    -0.25
    çİĩè¾¾
    -0.24
    $:
    -0.24
     die
    -0.24
    ardin
    -0.24
    POSITIVE LOGITS
    acier
    0.27
    wake
    0.25
    没å¿ħè¦ģ
    0.25
    åĺı
    0.24
    æĥ³åĬŀæ³ķ
    0.24
    olist
    0.24
    æĹłéľĢ
    0.24
    illy
    0.24
    ÙĦÙĬÙħ
    0.23
    没æľīå¿ħè¦ģ
    0.23
    Activations Density 2.811%

    No Known Activations