© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 14599
    Prev
    Next
    INDEX
    Explanations

    purpose and application

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

    The neuron fires on adverbial intensity or frequency modifiers (e.g. words like “often,” “most,” “especially,” “particularly”).

    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
     helicópter
    -1.23
     from
    -1.14
    bundet
    -1.13
     alemania
    -1.09
    OOTDTY
    -1.09
    ꒳
    -1.09
    ſſe
    -1.07
    腭
    -1.06
     Borussia
    -1.06
     each
    -1.06
    POSITIVE LOGITS
     used
    2.06
     for
    1.66
     use
    1.52
    สำหรับ
    1.15
    适用于
    1.09
     serve
    1.05
    P
    1.04
    遑
    1.02
     intended
    1.02
     utilized
    1.01
    Activations Density 0.129%

    No Known Activations