© 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. 72927
    Prev
    Next
    INDEX
    Explanations

    defining something as something else

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

    The neuron detects instances of the word “as” when it introduces a role, function, or classification (e.g. “X as Y”).

    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
     poichè
    -1.27
     affinch
    -1.26
     nových
    -1.19
    vereignty
    -1.19
    inerary
    -1.18
    盉
    -1.16
    などは
    -1.16
    iamo
    -1.16
    Fonto
    -1.13
     mediocre
    -1.12
    POSITIVE LOGITS
     a
    1.83
     an
    1.65
     our
    1.34
     delito
    1.33
     seule
    1.23
     juga
    1.23
     também
    1.23
     as
    1.20
     también
    1.20
     both
    1.19
    Activations Density 0.062%

    No Known Activations