© 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. 22-TRANSCODER-HP
    7. 111026
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    á»ģm
    -0.23
     +=↵
    -0.23
     pud
    -0.23
    boarding
    -0.23
    autoload
    -0.23
    eft
    -0.23
    мо
    -0.23
     lod
    -0.22
    “Yes
    -0.22
     bre
    -0.22
    POSITIVE LOGITS
    flight
    0.25
    ucks
    0.25
    å½±åĵįåĬĽçļĦ
    0.24
     flight
    0.24
     Tob
    0.23
    -flight
    0.23
    æĸ½
    0.23
    ç¿ħ
    0.23
    _HOT
    0.23
    æ°ijçĶ¨
    0.23
    Activations Density 0.000%

    No Known Activations

    This feature has no known activations.