© 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. 20-TRANSCODER-HP
    7. 108396
    Prev
    Next
    INDEX
    Explanations

    5

    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
    ditor
    -0.28
    åģļ强
    -0.25
    禺
    -0.25
     Debugger
    -0.24
    综åIJĪå¾ģ
    -0.23
    érer
    -0.23
     prospect
    -0.23
    çī©ä¸ļåħ¬åı¸
    -0.23
    lund
    -0.23
    驺
    -0.23
    POSITIVE LOGITS
    Rec
    0.33
     Rec
    0.32
    rec
    0.31
    9
    0.29
     rec
    0.26
    8
    0.25
    6
    0.25
    0
    0.25
     whatever
    0.25
     Against
    0.25
    Activations Density 18.731%

    No Known Activations