© 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. 34356
    Prev
    Next
    INDEX
    Explanations

    always

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

    The neuron selectively activates on the adverb “always,” marking habitual or recurring statements.

    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
     there
    -1.19
     parte
    -1.05
     is
    -1.05
     بعض
    -1.02
    जय
    -1.02
    OfType
    -1.01
     ditawarkan
    -1.01
    ↵↵↵↵↵↵↵↵
    -1.00
     потому
    -0.99
    ᑎ
    -0.99
    POSITIVE LOGITS
    ごとの
    1.31
    ほうが
    1.27
     škoda
    1.17
    になったら
    1.16
    '
    1.16
    larınız
    1.15
     always
    1.14
     vždy
    1.14
     thinking
    1.13
     wondered
    1.11
    Activations Density 0.018%

    No Known Activations