© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 105578
    Prev
    Next
    INDEX
    Explanations

    something unique

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

    The neuron flags promotional or descriptive phrases that emphasize novelty and difference—e.g. words like “different,” “unique,” “amazing,” “innovative,” “distinct,” “wonderful,” etc.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     wees
    -1.73
     SAME
    -1.63
     occuper
    -1.58
     goe
    -1.52
    侂
    -1.51
     ake
    -1.46
     vader
    -1.46
     rues
    -1.45
    indicates
    -1.44
     juf
    -1.43
    POSITIVE LOGITS
     that
    2.20
     or
    2.13
    <
    1.77
     when
    1.73
     at
    1.68
     bądź
    1.68
     before
    1.65
     ensure
    1.61
    ”
    1.59
     trying
    1.56
    Activations Density 0.021%

    No Known Activations