© 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. 62778
    Prev
    Next
    INDEX
    Explanations

    period

    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
    Negative Logits
    çīĩ
    -0.27
     manually
    -0.27
    ĴĮ
    -0.26
    ALTH
    -0.26
     Pyongyang
    -0.24
    è¶ķ
    -0.23
     liberties
    -0.23
    éĺĦ
    -0.23
    æīĭåĬ¨
    -0.23
    羯
    -0.23
    POSITIVE LOGITS
    component
    0.28
    reuse
    0.28
     moderation
    0.27
     Embed
    0.27
    ót
    0.26
    amac
    0.26
    çĨŁçŁ¥
    0.26
    èĺ¸
    0.25
    ow
    0.24
     component
    0.24
    Activations Density 7.152%

    No Known Activations