© 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. 25-TRANSCODER-HP
    7. 107572
    Prev
    Next
    INDEX
    Explanations

    Formal, technical language

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     imp
    -0.31
    orf
    -0.24
     vigor
    -0.23
    odom
    -0.23
     term
    -0.23
     link
    -0.23
     reign
    -0.23
    viewer
    -0.23
    çĤ
    -0.23
    dis
    -0.23
    POSITIVE LOGITS
    ements
    0.29
    ienne
    0.27
    pars
    0.27
    amientos
    0.26
    emens
    0.26
    åĻ¨
    0.26
    ateway
    0.26
    anter
    0.26
    ãģ¤
    0.26
    inent
    0.25
    Activations Density 0.128%

    No Known Activations