© 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. 65696
    Prev
    Next
    INDEX
    Explanations

    variable names followed by types

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

    The neuron fires on numeric literal tokens (e.g. standalone numbers or digits) in the text.

    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
     Muchas
    -1.10
    Whose
    -1.09
    来到了
    -1.05
    AutoModerator
    -1.05
     willst
    -1.04
     τὸν
    -1.04
    qas
    -1.01
     musst
    -1.00
     ਕੀ
    -0.99
     ως
    -0.99
    POSITIVE LOGITS
    leggio
    1.16
    ity
    1.12
     one
    1.09
    尼亚
    1.09
     that
    1.06
    ヶ月
    1.03
    舖
    1.02
     große
    1.01
     lägen
    1.01
     blistering
    1.00
    Activations Density 0.021%

    No Known Activations