© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 74756
    Prev
    Next
    INDEX
    Explanations

    relates to proximity or knowing someone

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

    The main thing this neuron does is detect the word “close” used in the phrase “close to” when referring to people who are emotionally or physically near someone.

    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
     if
    -1.61
     every
    -1.53
     because
    -1.48
     in
    -1.33
     there
    -1.32
     some
    -1.29
     what
    -1.27
     オリジナル
    -1.27
    Because
    -1.27
    𖡼
    -1.27
    POSITIVE LOGITS
    回到了
    1.31
     heureuse
    1.29
     ACTUALLY
    1.27
     attentively
    1.26
     revamped
    1.25
     réglable
    1.23
    especially
    1.23
    颊
    1.21
     ganzes
    1.20
     SUCH
    1.19
    Activations Density 0.014%

    No Known Activations