© 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. 33-TRANSCODER-HP
    7. 53468
    Prev
    Next
    INDEX
    Explanations

    text

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    æ²IJ
    -0.26
    æĻ¡
    -0.24
    furt
    -0.23
     Yugoslavia
    -0.23
     hep
    -0.23
    ä¸ŃåİŁ
    -0.23
    urn
    -0.23
    weit
    -0.22
    主åĬŀ
    -0.22
     dbo
    -0.22
    POSITIVE LOGITS
     text
    1.40
    æĸĩæľ¬
    1.30
    text
    1.20
     Text
    1.11
     textual
    1.06
     ÑĤекÑģÑĤ
    1.04
     texto
    1.00
    (text
    1.00
    Text
    0.99
    .text
    0.99
    Activations Density 0.138%

    No Known Activations