© 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. 73009
    Prev
    Next
    INDEX
    Explanations

    configuration files and code

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

    The neuron primarily fires on standalone “.” tokens (dots), such as the “.” in commands like COPY . . and on lines beginning with “.” in assembly or markup directives.

    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
    :
    -2.23
    .”
    -2.22
    方針
    -2.11
    вня
    -2.08
     What
    -1.97
    ☀
    -1.95
     Only
    -1.95
    恐れ
    -1.94
    もしか
    -1.93
     作り方
    -1.92
    POSITIVE LOGITS
    雠
    2.64
    臯
    2.50
    粜
    2.45
     trám
    2.38
     flourished
    2.34
     thrived
    2.23
    訁
    2.19
    {
    2.17
     cóc
    2.17
     jeopardi
    2.17
    Activations Density 0.031%

    No Known Activations