© 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. 128910
    Prev
    Next
    INDEX
    Explanations

    Code and literacy

    np_max-act · gemini-2.0-flash
    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
    avel
    -0.32
    apper
    -0.28
     glacier
    -0.27
    apel
    -0.27
    zk
    -0.26
     slump
    -0.25
     dragons
    -0.25
     kort
    -0.24
     glaciers
    -0.23
    plx
    -0.23
    POSITIVE LOGITS
    urally
    0.29
    èѦ
    0.28
    ecn
    0.28
    STS
    0.28
    ImageContext
    0.25
    èĭ¦
    0.25
    esome
    0.25
    ime
    0.25
    rib
    0.25
    ç³Ł
    0.25
    Activations Density 0.002%

    No Known Activations