© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 19-TRANSCODER-HP
    7. 6620
    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
    aptor
    -0.26
    nia
    -0.25
     LES
    -0.23
     improv
    -0.23
    èĩªçIJĨ
    -0.23
    ackbar
    -0.23
    -scrollbar
    -0.23
    åĪ¿
    -0.23
    ioni
    -0.23
    稼
    -0.23
    POSITIVE LOGITS
    èħĭ
    0.31
    çĵ¦
    0.28
    :type
    0.26
    as
    0.25
    uche
    0.25
    office
    0.25
    apel
    0.24
    èħIJ
    0.24
    æ¡Ī
    0.24
    é¦ĸ
    0.24
    Activations Density 0.580%

    No Known Activations

    This feature has no known activations.