© 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. 19-TRANSCODER-HP
    7. 8036
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_19.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.19.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     stal
    -0.27
     Xm
    -0.27
    avo
    -0.26
    esimal
    -0.25
    çľĭè¿ĩ
    -0.25
    iesta
    -0.24
    maker
    -0.24
    ests
    -0.24
    çľ¼çļ®
    -0.24
    aldo
    -0.24
    POSITIVE LOGITS
    than
    0.28
    >y
    0.27
    æĵį
    0.24
    eration
    0.24
    mys
    0.24
    åįķçĭ¬
    0.24
    ordinate
    0.24
    ention
    0.24
     than
    0.23
    bsolute
    0.23
    Activations Density 0.093%

    No Known Activations

    This feature has no known activations.