© 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. Gemma-2-2B
    5. 0-GEMMASCOPE-TRANSCODER-16K
    7. 4341
    Prev
    Next
    INDEX
    Explanations

    the word "Corporation"

    oai_token-act-pair · gemini-2.0-flash

    corporation

    np_max-act · gemini-2.0-flash

    Corporation

    np_max-act-logits · gpt-5-miniTriggered by @chenshw0109
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_0/width_16k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.0.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     
    -1.63
    ↵↵
    -1.54
     the
    -1.50
     a
    -1.50
     (
    -1.49
    ,
    -1.48
    .
    -1.43
     in
    -1.41
     to
    -1.41
    ↵
    -1.26
    POSITIVE LOGITS
     Efq
    2.36
     myſelf
    2.23
     ་་
    2.22
     itſelf
    2.22
     ―――――
    2.16
     doubtnut
    2.09
     $_"
    1.95
    ſelf
    1.95
    IUrlHelper
    1.93
     Jefus
    1.92
    Activations Density 1.848%

    No Known Activations