© 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. 31-TRANSCODER-HP
    7. 19489
    Prev
    Next
    INDEX
    Explanations

    change, influence, growth

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     not
    -0.28
    —
    -0.26
    åĻ¨
    -0.25
    —if
    -0.25
    ”
    -0.25
     NOT
    -0.24
    /download
    -0.24
    öt
    -0.24
    .SEVERE
    -0.24
    ĥ
    -0.24
    POSITIVE LOGITS
     Stage
    0.32
     Begins
    0.32
     Goes
    0.32
     Starts
    0.32
     Makes
    0.31
     Phase
    0.30
     Means
    0.30
     Problems
    0.30
     Edge
    0.30
     happens
    0.30
    Activations Density 3.770%

    No Known Activations