© 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. 98971
    Prev
    Next
    INDEX
    Explanations

    being adjectives

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

    The neuron activates strongly on descriptive adjectives (and related adverbs) that characterize qualities (e.g. imaginative, intentional, innovative, expressive, confusing, vague, specific).

    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
     that
    -1.77
     destinée
    -1.40
     aidé
    -1.38
     confé
    -1.34
     voyez
    -1.31
     impegn
    -1.30
     fréqu
    -1.30
    瀕
    -1.30
     Quando
    -1.30
     modelleri
    -1.26
    POSITIVE LOGITS
    6
    2.19
     in
    2.09
    7
    2.03
    3
    1.81
    8
    1.75
    4
    1.74
    9
    1.73
    2
    1.70
    5
    1.62
    1
    1.52
    Activations Density 0.053%

    No Known Activations