© 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. 87518
    Prev
    Next
    INDEX
    Explanations

    code snippets

    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
    imax
    -0.30
    æ¥Ķ
    -0.28
    身æĹģ
    -0.28
     ÄijÃłn
    -0.28
    åįģåŃĹ
    -0.27
    ç´¯äºĨ
    -0.26
    æ¶İ
    -0.25
    diff
    -0.24
    uche
    -0.24
    æ¯ĶèµĽä¸Ń
    -0.24
    POSITIVE LOGITS
     hope
    0.28
    å¸ĮæľĽ
    0.27
    backs
    0.27
    åĵ²
    0.26
    лаг
    0.26
     trans
    0.25
    çIJ¯
    0.25
    ati
    0.24
     cig
    0.24
    building
    0.24
    Activations Density 0.013%

    No Known Activations