© 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. 65733
    Prev
    Next
    INDEX
    Explanations

    coming out of

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

    The neuron fires on verbs of movement or change of state—words like “leave,” “come,” “emerge,” or “back” that signal an action or transition.

    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
     to
    -1.61
     już
    -1.16
     SQLException
    -1.10
    /**/
    -1.01
    obut
    -1.00
    ditionally
    -0.99
     Halen
    -0.99
     già
    -0.97
    ческой
    -0.96
     أثناء
    -0.96
    POSITIVE LOGITS
     both
    1.30
     satisfied
    1.13
    excellents
    1.09
     vanquished
    1.02
    culosis
    1.01
     feeling
    1.01
     concluding
    1.01
     результа
    1.00
    ̍t
    1.00
     cueva
    0.99
    Activations Density 0.014%

    No Known Activations