© 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. 95360
    Prev
    Next
    INDEX
    Explanations

    Aug

    np_max-act-logits · gemini-2.0-flash
    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
    éĹ®ä¸ĸ
    -0.28
    getRoot
    -0.27
    .getRoot
    -0.27
    lin
    -0.27
     Bin
    -0.26
    tron
    -0.26
    /ref
    -0.26
    Logic
    -0.25
    inus
    -0.25
     *(*
    -0.24
    POSITIVE LOGITS
    éľĩåĬ¨
    0.31
    kop
    0.27
     redirects
    0.27
     preempt
    0.27
    æĢģ
    0.26
    åĵģ
    0.25
    oc
    0.24
    ures
    0.24
    ificates
    0.23
     ostat
    0.23
    Activations Density 0.882%

    No Known Activations