© 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. 35-TRANSCODER-HP
    7. 39363
    Prev
    Next
    INDEX
    Explanations

    Scientific/technical introductions

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_35.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.35.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     ============================================================================↵
    -0.30
    agos
    -0.27
    Mill
    -0.25
    aget
    -0.24
    æĥħåĨµè¿Ľè¡Į
    -0.24
    croft
    -0.24
    geois
    -0.24
    dì
    -0.24
    venge
    -0.24
    ationally
    -0.24
    POSITIVE LOGITS
     Stub
    0.31
    uell
    0.27
    utions
    0.26
     hum
    0.25
    .testing
    0.25
    ë¦Ń
    0.25
    ÑģÑĭ
    0.24
     churn
    0.24
     nib
    0.23
    Stub
    0.23
    Activations Density 3.944%

    No Known Activations