© 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. 17816
    Prev
    Next
    INDEX
    Explanations

    Human attributes

    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
    elf
    -0.27
    éĢļ讯åijĺ
    -0.26
    nte
    -0.26
    eners
    -0.26
    ense
    -0.26
     addons
    -0.25
    ItemAt
    -0.25
    ode
    -0.24
    æĹ¥è®¯
    -0.24
    å¢ĥå¤ĸ
    -0.23
    POSITIVE LOGITS
     lives
    0.94
     minds
    0.82
     heads
    0.78
     bodies
    0.75
     selves
    0.75
     souls
    0.75
     hearts
    0.73
     voices
    0.73
     careers
    0.72
     personalities
    0.69
    Activations Density 73.315%

    No Known Activations