© 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. 96025
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    agara
    -0.27
    Dirs
    -0.26
    fresh
    -0.25
    åĩ½
    -0.25
    è¿Ń
    -0.25
    èķ´
    -0.25
    æĽĿ
    -0.25
     accumulated
    -0.25
     dips
    -0.24
    imentary
    -0.24
    POSITIVE LOGITS
    åĮĹæŀģ
    0.27
    ç®Ń
    0.25
     Earn
    0.25
    第ä¸ĢæŃ¥
    0.24
    缴æµģ
    0.24
     roleName
    0.23
    ÑĤÑĥ
    0.23
     provision
    0.23
     dressing
    0.23
    ONO
    0.23
    Activations Density 0.882%

    No Known Activations

    This feature has no known activations.