© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 44230
    Prev
    Next
    INDEX
    Explanations

    family members

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

    The neuron detects kinship terms—words naming family relationships (e.g., mother, father, parents, daughter, son, sister).

    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
    πισ
    -1.20
     izdel
    -1.20
     it
    -1.17
     you
    -1.15
     usar
    -1.12
     conseguir
    -1.09
     trei
    -1.08
    slowly
    -1.05
     viejo
    -1.04
     any
    -1.03
    POSITIVE LOGITS
     staffel
    1.16
     Seem
    1.08
     laissant
    1.07
    そうに
    1.06
     effekt
    1.05
     nabo
    1.05
     kurikulum
    1.05
     karte
    1.05
    так
    1.03
     pome
    1.02
    Activations Density 0.083%

    No Known Activations