© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 22-TRANSCODER-HP
    7. 130828
    Prev
    Next
    INDEX
    Explanations

    Code, names, software

    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
    Not in Any Lists

    No Comments

    Negative Logits
    hör
    -0.25
    idend
    -0.25
    chn
    -0.24
    ipay
    -0.24
    éĢıéģİ
    -0.24
    ampa
    -0.24
    amenti
    -0.24
    nock
    -0.24
    ndo
    -0.23
     thá»ķ
    -0.23
    POSITIVE LOGITS
    ources
    0.28
    uns
    0.25
    ello
    0.25
    å°±ç»Ļ大家
    0.24
    å¿«ä¸ī
    0.24
    sess
    0.23
    æĸĹ
    0.23
     grabs
    0.23
     cares
    0.23
    vens
    0.23
    Activations Density 0.028%

    No Known Activations