© 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. 14-TRANSCODER-HP
    7. 41680
    Prev
    Next
    INDEX
    Explanations

    print

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_14.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.14.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    æĬĬ她
    -0.28
    exact
    -0.27
    licer
    -0.27
    Utf
    -0.26
    详ç»ĨäºĨ解
    -0.26
     exact
    -0.26
    :image
    -0.25
    ä½ĥ
    -0.25
    rets
    -0.25
    heed
    -0.24
    POSITIVE LOGITS
    .band
    0.28
    memberOf
    0.26
    rais
    0.25
    .dtd
    0.24
     formulas
    0.24
    iado
    0.24
    iscal
    0.24
    earing
    0.23
    -band
    0.23
    iad
    0.23
    Activations Density 0.378%

    No Known Activations