© 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. 20-TRANSCODER-HP
    7. 141335
    Prev
    Next
    INDEX
    Explanations

    small amounts

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     TestBed
    -0.27
    losures
    -0.27
    trs
    -0.26
    cies
    -0.24
    olas
    -0.24
    xes
    -0.24
     trials
    -0.24
     stranded
    -0.23
     duties
    -0.23
    /inet
    -0.23
    POSITIVE LOGITS
    ç¼ĵæħ¢
    0.43
     slow
    0.37
    æħ¢
    0.35
     continuously
    0.34
     gradual
    0.34
    åĿĩåĮĢ
    0.33
     slowly
    0.33
    åħ¨ç¨ĭ
    0.33
    æŁĶåĴĮ
    0.33
    稳æŃ¥æİ¨è¿Ľ
    0.32
    Activations Density 75.265%

    No Known Activations