© 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. 72500
    Prev
    Next
    INDEX
    Explanations

    processes and their results

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

    The neuron fires strongly on formal, technical process words—especially nouns built with “-ation,” “-ator,” etc. (e.g. cremation/cremated, sterilization/sterilizing, incinerator/incineration).

    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
    ങ്ങള
    -1.31
    ໌
    -1.30
     정보
    -1.30
    ଝ
    -1.28
     发型
    -1.27
    Voltaje
    -1.25
    licitations
    -1.23
    されていない
    -1.23
     rots
    -1.23
    גי
    -1.23
    POSITIVE LOGITS
     “
    1.20
     没有
    1.09
     flexible
    1.05
    i
    1.05
    }
    1.02
    berfläche
    1.00
     option
    1.00
     simple
    0.99
     Florida
    0.98
     usually
    0.98
    Activations Density 0.002%

    No Known Activations