© 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. 22-GEMMASCOPE-RES-131K
    7. 15192
    Prev
    Next
    INDEX
    Explanations

    brother or sister

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

    The neuron responds whenever it sees a numeric token (multi-digit numbers).

    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
     contenedores
    -0.96
     unchanging
    -0.90
    Eid
    -0.90
     stattfinden
    -0.88
    ʾ
    -0.88
    Abit
    -0.88
     范
    -0.86
    只需要
    -0.84
    tedy
    -0.83
     reverting
    -0.83
    POSITIVE LOGITS
     brother
    1.14
     sister
    1.13
    brother
    0.97
    𝟒
    0.97
     IBOutlet
    0.96
     sisters
    0.92
     macchina
    0.92
    eload
    0.92
    sister
    0.91
     brothers
    0.91
    Activations Density 0.002%

    No Known Activations