© 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. 22-GEMMASCOPE-RES-131K
    7. 127832
    Prev
    Next
    INDEX
    Explanations

    The neuron responds to strongly evaluative adjectives or adverbs (words that convey a speaker’s judgment or emphatic tone).

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408

    words like insidious, eyebrow, mood

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    Negative Logits
     liberi
    -1.01
     呢
    -0.99
    such
    -0.99
    ようです
    -0.97
    あー
    -0.96
     resembling
    -0.96
    顷
    -0.95
    轶
    -0.94
    قامت
    -0.94
     такой
    -0.93
    POSITIVE LOGITS
    uating
    1.07
     NACIONAL
    1.05
    opropyl
    1.04
    ichord
    1.03
    ALLENGES
    1.03
    ınt
    1.02
     öf
    1.01
    vereign
    0.98
    ocyan
    0.97
    cketed
    0.96
    Activations Density 0.037%

    No Known Activations