© 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. 108400
    Prev
    Next
    INDEX
    Explanations

    Non-English words

    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
    ½æķ°
    -0.30
    oppel
    -0.27
    .bp
    -0.27
    ignet
    -0.27
    beits
    -0.26
     begr
    -0.26
    inesis
    -0.25
     quantify
    -0.25
    icolon
    -0.25
    éĩįçĤ¹é¢ĨåŁŁ
    -0.25
    POSITIVE LOGITS
    åĬ¡
    0.30
     Dress
    0.27
     dressing
    0.26
    ç¶ļãģį
    0.26
     dealing
    0.25
    ghi
    0.25
    _sales
    0.25
    Ñij
    0.25
    害ç¾ŀ
    0.25
    éģį
    0.24
    Activations Density 0.909%

    No Known Activations