© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 31-TRANSCODER-HP
    7. 18107
    Prev
    Next
    INDEX
    Explanations

    Related to "take"

    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
    FOUND
    -0.29
    akra
    -0.28
    holds
    -0.26
    allet
    -0.26
     imag
    -0.25
    ç»Ļåĩº
    -0.25
    åĽłä¸ºåľ¨
    -0.24
     constructs
    -0.24
    clare
    -0.24
    Found
    -0.24
    POSITIVE LOGITS
     taken
    0.52
    taken
    0.45
    åıĸ
    0.44
    åıĸãĤĬ
    0.39
     took
    0.34
     Taken
    0.34
     take
    0.34
     taking
    0.33
    _taken
    0.33
     Take
    0.33
    Activations Density 1.808%

    No Known Activations