© 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. 39456
    Prev
    Next
    INDEX
    Explanations

    Louis

    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
     mutating
    -0.30
    “(
    -0.27
     french
    -0.26
     offending
    -0.26
    æĥħåĨµè¿Ľè¡Į
    -0.26
     mutate
    -0.25
     cooperating
    -0.25
    åįĬ个
    -0.24
    ä¼ĹæīĢ
    -0.24
    æĬķæľº
    -0.24
    POSITIVE LOGITS
    e
    0.29
     Raz
    0.27
    inho
    0.27
    ánÃŃ
    0.26
     character
    0.26
    ienda
    0.26
    i
    0.26
    anje
    0.25
    Ãł
    0.25
     Vas
    0.25
    Activations Density 0.903%

    No Known Activations