© 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. 31-TRANSCODER-HP
    7. 18739
    Prev
    Next
    INDEX
    Explanations

    say "number"

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ½
    -0.28
    ä¸ĢåĪĨ
    -0.28
    ¼
    -0.25
    à¹Ģลย
    -0.24
    ä¸Ģåij¨
    -0.24
    è¿ĺ羣
    -0.24
     Fra
    -0.23
    .fp
    -0.23
     [.
    -0.23
    á¹£
    -0.23
    POSITIVE LOGITS
    åħŃåįģ
    0.63
    äºĶåįģ
    0.60
     fifty
    0.58
    ä¸īåįģ
    0.58
     sixty
    0.57
     eighty
    0.56
    åĽĽåįģ
    0.54
    ä¸ĥåįģ
    0.54
     seventy
    0.54
    åħ«åįģ
    0.54
    Activations Density 21.181%

    No Known Activations