Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Gemma-2-2B
    3. 2-GEMMASCOPE-TRANSCODER-16K
    4. 14263
    Prev
    Next
    INDEX
    Explanations

    This neuron seems to be a hodgepodge, responding to mentions of "compiz", the French word "tu/je," legal/political terms, and other random words, so it is hard to give one single explanation.

    oai_token-act-pair · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope/layer_2/width_16k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    но
    -0.91
    spare
    -0.88
     myſelf
    -0.86
     spare
    -0.86
     faſt
    -0.85
     Theſe
    -0.79
     iſt
    -0.78
     ―――――
    -0.78
    LookAnd
    -0.76
     purpoſe
    -0.76
    POSITIVE LOGITS
    NotNil
    0.47
    {}
    0.46
    valently
    0.43
    ${\
    0.42
     
    0.42
    aba
    0.42
     Greg
    0.41
    ?
    0.40
    .”
    0.40
    “
    0.40
    Activations Density 4.092%

    No Known Activations