© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 54178
    Prev
    Next
    INDEX
    Explanations

    I, It, president

    np_acts-logits-general · gemini-2.5-flash-lite

    This neuron fires on the initial content words of answer segments (the first token after each “A:”), marking the start of answer utterances.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    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
     also
    -2.73
     también
    -2.20
     também
    -2.06
     גם
    -2.02
     также
    -1.97
    Also
    -1.95
     ALSO
    -1.93
     także
    -1.88
     then
    -1.86
    also
    -1.79
    POSITIVE LOGITS
    Depends
    1.86
     Depends
    1.69
     depends
    1.66
    depends
    1.53
     actually
    1.49
    当然是
    1.49
     certainly
    1.44
    Answered
    1.42
     definitely
    1.41
     varies
    1.32
    Activations Density 0.068%

    No Known Activations