© 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. 30-TRANSCODER-HP
    7. 95619
    Prev
    Next
    INDEX
    Explanations

    dangerous substances

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    axon
    -0.28
    alli
    -0.27
    ç®Ĺæ³ķ
    -0.27
     Cycling
    -0.26
    æīĵæĸŃ
    -0.25
    mino
    -0.25
    _algorithm
    -0.25
     algorithm
    -0.25
     forg
    -0.24
    oliberal
    -0.24
    POSITIVE LOGITS
    åį±éĻ©
    0.35
     substances
    0.29
     substance
    0.29
    åį±éļª
    0.28
     hazardous
    0.28
    danger
    0.28
     materials
    0.27
     material
    0.27
    åºŁçī©
    0.27
     clutter
    0.27
    Activations Density 0.886%

    No Known Activations