© 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. 113927
    Prev
    Next
    INDEX
    Explanations

    programming keywords and actions

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

    variable names or identifiers in programming code that are written in camelCase format.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron fires most strongly on programming identifier tokens (variable, function, or constant names) rather than on ordinary language words.

    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
     gedurende
    -1.61
     with
    -1.55
     kdo
    -1.53
     our
    -1.48
     ciertamente
    -1.46
    蝽
    -1.42
     bedoeld
    -1.41
     to
    -1.40
     verdaderamente
    -1.40
    UALA
    -1.38
    POSITIVE LOGITS
    6
    2.16
    7
    1.97
    8
    1.82
    3
    1.80
    Because
    1.59
    They
    1.55
    ここで
    1.55
    It
    1.53
     их
    1.52
    5
    1.49
    Activations Density 0.134%

    No Known Activations