© 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. 18052
    Prev
    Next
    INDEX
    Explanations

    source code, documentation

    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
     about
    -0.28
     fine
    -0.25
     Bowen
    -0.25
    inet
    -0.25
    iameter
    -0.24
     Feed
    -0.24
     uncont
    -0.24
    eer
    -0.24
    å¤ĩ
    -0.23
     portal
    -0.23
    POSITIVE LOGITS
    ç»ĻåĩºäºĨ
    0.28
    uttle
    0.28
    .nano
    0.27
    rous
    0.26
    åºĦ严
    0.26
    çģµéŃĤ
    0.25
    [--
    0.25
    gaard
    0.25
    adh
    0.25
    å¢ħ
    0.25
    Activations Density 0.002%

    No Known Activations