© 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. 30-TRANSCODER-HP
    7. 96801
    Prev
    Next
    INDEX
    Explanations

    technical documentation

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ursed
    -0.32
     Trudeau
    -0.32
     Canadians
    -0.31
    åĬłæĭ¿å¤§
    -0.31
    ziel
    -0.29
    çī©
    -0.27
     canada
    -0.27
    ç®Ĺ
    -0.27
    漫
    -0.26
    äºĮ级
    -0.26
    POSITIVE LOGITS
     skinny
    0.27
    bers
    0.26
    erge
    0.26
    eria
    0.25
    rape
    0.25
    (k
    0.25
     leg
    0.25
     UIL
    0.24
    é£İåħī
    0.24
    kn
    0.24
    Activations Density 0.001%

    No Known Activations