© 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. 73562
    Prev
    Next
    INDEX
    Explanations

    name variations and recognition

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

    The neuron fires on technical terms and verbs around string‐matching or name‐normalization tasks—especially “name,” “match,” “variations,” “using,” etc.—i.e. it detects when the text is about matching or normalizing names/strings in code.

    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
    Élet
    -0.98
     eignet
    -0.92
     indepen
    -0.90
    Unnamed
    -0.89
     rama
    -0.89
    ilities
    -0.89
    sikal
    -0.89
     bağı
    -0.87
    贡献
    -0.86
    thenReturn
    -0.85
    POSITIVE LOGITS
    ឈ
    1.03
     recognised
    1.01
     recognise
    0.98
     leur
    0.95
    Причины
    0.93
     variations
    0.92
     recognition
    0.92
     uved
    0.90
     recognize
    0.90
     alasan
    0.88
    Activations Density 0.122%

    No Known Activations