© 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. 10-GEMMASCOPE-RES-131K
    7. 105876
    Prev
    Next
    INDEX
    Explanations

    Stylus pen, plunger, sourdough, Stylonurella

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

    The neuron reliably fires on short content words built around a consonant–liquid–consonant cluster (e.g. stylus, stylonurella, style, stylize, stiletos), essentially picking out “stl/str/pl”-type consonant–liquid sequences in tokens.

    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
    ve
    -3.17
    ).
    -2.67
    铼
    -2.03
    *:
    -2.03
    et
    -2.02
    amoan
    -2.00
     hemel
    -1.99
    켭
    -1.95
     består
    -1.95
    </em>
    -1.93
    POSITIVE LOGITS
    ꪔ
    2.63
    たぶん
    2.22
    ・・・・・・・・
    2.16
     poziom
    2.05
    Focusing
    2.05
     soulign
    2.02
     agarre
    1.99
    ὺς
    1.96
    久し
    1.95
    ?”
    1.95
    Activations Density 0.004%

    No Known Activations