© 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. 22-GEMMASCOPE-RES-131K
    7. 36503
    Prev
    Next
    INDEX
    Explanations

    family and myself

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

    The neuron activates on personal, first‐person references—pronouns like “my” or “I” and names of the author’s close relations or collaborators.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     so
    -0.93
    گ
    -0.91
    CastException
    -0.87
     riferimento
    -0.86
    についての
    -0.85
     pregunta
    -0.85
    を
    -0.85
    OK
    -0.85
    判断
    -0.83
    ‌
    -0.82
    POSITIVE LOGITS
     and
    1.16
     myself
    1.15
    和我
    1.10
     me
    1.06
     μαζί
    1.03
     мной
    1.00
     I
    0.99
     together
    0.97
     împreună
    0.95
     my
    0.90
    Activations Density 0.030%

    No Known Activations