© 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. 22-TRANSCODER-HP
    7. 98143
    Prev
    Next
    INDEX
    Explanations

    Nonsensical/mixed context text

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    为åĩĨ
    -0.27
    street
    -0.26
    åĭ
    -0.25
    æŃ¤åŁºç¡Ģä¸Ĭ
    -0.24
    ино
    -0.24
    请çĤ¹åĩ»
    -0.23
     getRandom
    -0.23
     createElement
    -0.23
    argout
    -0.23
    ä¸įå¾Ĺä¸į
    -0.23
    POSITIVE LOGITS
     researchers
    0.32
     proponents
    0.26
    ĵ
    0.26
    rig
    0.25
    /gcc
    0.25
     rig
    0.25
    aklı
    0.24
    æ²®
    0.24
    带æĿ¥çļĦ
    0.24
     taxpayers
    0.24
    Activations Density 0.003%

    No Known Activations