© 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. 19932
    Prev
    Next
    INDEX
    Explanations

    programming terms

    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
    -earth
    -0.28
     station
    -0.28
    bed
    -0.27
     Lay
    -0.26
    大èĩªçĦ¶
    -0.25
     stations
    -0.24
     laid
    -0.24
    æĶ¯æĴij
    -0.24
    èī¯å¥½
    -0.24
     fair
    -0.24
    POSITIVE LOGITS
    zas
    0.27
    rk
    0.27
    æĤ¬æĮĤ
    0.27
    ivas
    0.26
    illos
    0.26
    æĮ¥åıij
    0.26
    stim
    0.25
    umd
    0.24
     Stim
    0.24
     commit
    0.24
    Activations Density 0.960%

    No Known Activations