© 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. 18696
    Prev
    Next
    INDEX
    Explanations

    Technical/programming context

    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
     release
    -0.31
    (es
    -0.29
    frag
    -0.29
    éĩĬæĶ¾
    -0.28
    lear
    -0.28
    -hook
    -0.28
    imir
    -0.27
    -release
    -0.27
    _release
    -0.27
     Sec
    -0.27
    POSITIVE LOGITS
    pare
    0.27
     sop
    0.26
    çªĹåı£
    0.26
     decals
    0.26
    >window
    0.26
     kaz
    0.26
     payday
    0.25
    getWindow
    0.25
     pare
    0.25
    sizeof
    0.25
    Activations Density 6.439%

    No Known Activations