© 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. 98501
    Prev
    Next
    INDEX
    Explanations

    parents, separation, and children

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

    This neuron responds to mentions of a person’s father or dad and related childhood/age context (e.g. “father,” “dad,” “when she was X months old”).

    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
     individuelles
    -1.05
     to
    -1.00
    うどん
    -0.97
    摆脱
    -0.96
    了出去
    -0.96
    zzlies
    -0.94
    previously
    -0.93
    却又
    -0.92
    整个人
    -0.91
    捅
    -0.91
    POSITIVE LOGITS
     father
    1.37
     parents
    1.28
     family
    1.26
    dö
    1.26
     children
    1.25
     mother
    1.22
     leaving
    1.20
    Ŝ
    1.17
    лік
    1.13
     partidas
    1.10
    Activations Density 0.024%

    No Known Activations