© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 99003
    Prev
    Next
    INDEX
    Explanations

    real nice, bad, ugly, sharp, gone

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

    The neuron fires on colloquial intensifiers—single words like “real,” “nice,” “bad,” “ugly,” “quick,” or “sharp” used to amplify or emphasize a description.

    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
    肮
    -1.63
     présenter
    -1.52
     evenement
    -1.52
    遽
    -1.52
     éven
    -1.51
    ript
    -1.48
     nuevos
    -1.46
    嬸
    -1.45
     MULTIPLE
    -1.44
    pueden
    -1.44
    POSITIVE LOGITS
    ),
    1.86
    ."
    1.86
    l
    1.80
    ")
    1.77
    ]
    1.76
    d
    1.67
     only
    1.62
    。
    1.61
    r
    1.60
    c
    1.59
    Activations Density 0.015%

    No Known Activations