© 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. 10-GEMMASCOPE-RES-131K
    7. 105492
    Prev
    Next
    INDEX
    Explanations

    I, my, will, we

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

    The neuron fires strongly on the “Q:” question prompt lines, i.e. it detects the start of each question in the Q / A format.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     Hinblick
    -2.09
     like
    -1.95
     participaron
    -1.93
    ㉨
    -1.92
     keep
    -1.84
    ”،
    -1.81
     what
    -1.81
     aches
    -1.79
     deportivas
    -1.77
     avoid
    -1.77
    POSITIVE LOGITS
    "
    2.22
    decks
    2.09
     itali
    2.08
    獁
    2.06
     We
    2.00
    赑
    2.00
    ro
    1.98
    俢
    1.96
     hain
    1.95
    ‘
    1.93
    Activations Density 0.072%

    No Known Activations