© 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. 33616
    Prev
    Next
    INDEX
    Explanations

    person names and titles

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

    The neuron fires strongly on capitalized words that are part of people’s names (proper names).

    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
     geschickt
    -0.97
    ierz
    -0.96
    いていた
    -0.95
     rahat
    -0.91
     quatrième
    -0.91
    ఱ
    -0.90
     Chriftian
    -0.90
    いてる
    -0.88
    せっかく
    -0.88
     hauptsächlich
    -0.87
    POSITIVE LOGITS
     and
    1.63
     every
    1.26
     but
    1.23
     any
    1.15
     becomes
    1.13
    ylvan
    1.09
     also
    1.07
     your
    1.07
     both
    1.04
     always
    1.02
    Activations Density 0.001%

    No Known Activations