© 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. 15626
    Prev
    Next
    INDEX
    Explanations

    the word "source" and variations of "light"

    oai_token-act-pair · gemini-2.0-flash

    source

    np_max-act · gemini-2.0-flash
    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
    astéro
    -0.73
    RTEE
    -0.62
     Nazare
    -0.60
     
    -0.58
    ِّف
    -0.55
     _
    -0.54
     Switches
    -0.54
    DataMember
    -0.53
    :✨
    -0.53
     ...
    -0.53
    POSITIVE LOGITS
    <bos>
    1.19
     Monfieur
    0.96
     pleaſure
    0.93
     Efq
    0.85
     Jefus
    0.85
     purpoſe
    0.83
     leaſt
    0.83
     myſelf
    0.82
     BoxDecoration
    0.80
     raiſ
    0.79
    Activations Density 0.595%

    No Known Activations