© 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. 106936
    Prev
    Next
    INDEX
    Explanations

    mathematical equations

    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
    ahi
    -0.28
     XCT
    -0.26
     asked
    -0.25
    terminal
    -0.25
    ordered
    -0.24
    æĭĽçĶŁ
    -0.24
     ordered
    -0.24
     exact
    -0.24
    ibility
    -0.24
    Maker
    -0.24
    POSITIVE LOGITS
    >Password
    0.28
     Guerrero
    0.25
    åıijå±ķä¸Ń
    0.24
    ä¸Ĭåįĩ
    0.24
    çĦ¶
    0.23
    çļĦåķĨåĵģ
    0.23
     Guth
    0.23
    lew
    0.23
    ooth
    0.23
    çļĦåİĭåĬĽ
    0.23
    Activations Density 21.939%

    No Known Activations