© 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. 18127
    Prev
    Next
    INDEX
    Explanations

    to

    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
    gee
    -0.29
    gl
    -0.28
    é¢Ī
    -0.26
    çIJ¶
    -0.26
    peror
    -0.26
    åĬĪ
    -0.25
    heure
    -0.25
    åijĽ
    -0.25
    沿线
    -0.25
    éĵħ
    -0.24
    POSITIVE LOGITS
     bulky
    0.29
    artz
    0.29
     mamm
    0.26
     wet
    0.26
    ãģĹãģķ
    0.26
     saturated
    0.25
    ä¸Ģéģį
    0.25
     humanity
    0.25
     flagged
    0.25
    å·´æĸ¯
    0.25
    Activations Density 2.405%

    No Known Activations