© 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. 18006
    Prev
    Next
    INDEX
    Explanations

    well

    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
    Negative Logits
     stake
    -0.33
    _multiple
    -0.32
     multiple
    -0.29
    multiple
    -0.28
    å¤ļéĩį
    -0.26
    .tpl
    -0.25
    /from
    -0.25
    acious
    -0.25
     Multiple
    -0.25
    BlockSize
    -0.24
    POSITIVE LOGITS
     owned
    0.31
     Owned
    0.29
     borne
    0.28
    碰
    0.28
    æľīä¸Ģä½į
    0.28
     supported
    0.27
    brew
    0.27
    omed
    0.26
    upported
    0.26
     recognizing
    0.26
    Activations Density 2.150%

    No Known Activations