© 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. 18-TRANSCODER-HP
    7. 88164
    Prev
    Next
    INDEX
    Explanations

    Digits

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    å¿ĥè¡Ģ
    -0.32
     pitches
    -0.27
     immunity
    -0.27
    çªĿ
    -0.27
     pitch
    -0.26
     ult
    -0.26
     resp
    -0.26
    æ²ģ
    -0.26
    úb
    -0.25
    Former
    -0.25
    POSITIVE LOGITS
    ragen
    0.28
    placement
    0.27
    React
    0.25
    kiye
    0.25
    icut
    0.24
    angep
    0.24
    å¾®åŀĭ
    0.23
    ields
    0.23
    placements
    0.23
     React
    0.23
    Activations Density 13.561%

    No Known Activations