© 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. 88141
    Prev
    Next
    INDEX
    Explanations

    until

    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
    åĴ»
    -0.26
     away
    -0.26
     cho
    -0.26
    eros
    -0.25
    iding
    -0.25
    iedo
    -0.25
    \d
    -0.25
    çļĦéĢīæĭ©
    -0.25
    ceso
    -0.24
    ded
    -0.24
    POSITIVE LOGITS
    burgh
    0.25
    absolute
    0.24
    HostException
    0.24
     engr
    0.24
    ibre
    0.23
    åĪĩå®ŀ
    0.23
    深交
    0.23
     Absolute
    0.23
    -xl
    0.23
    ç§ij
    0.23
    Activations Density 15.326%

    No Known Activations