© 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. 34-GEMMASCOPE-RES-131K
    7. 47092
    Prev
    Next
    INDEX
    Explanations

    code snippets and structure

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

    The neuron strongly activates on programming‐style tokens—especially numeric literals and code identifiers (e.g. version numbers, class/file names, CamelCase identifiers) rather than normal prose words.

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
    lüssel
    -0.80
     Findings
    -0.73
    Dieu
    -0.68
    ופת
    -0.68
    进行了
    -0.67
     Voy
    -0.65
    Findings
    -0.65
     approvals
    -0.65
    下車
    -0.64
    ustus
    -0.63
    POSITIVE LOGITS
    otte
    0.84
    external
    0.76
     Verwaltungs
    0.71
    maps
    0.69
    ΡΑ
    0.69
    theme
    0.66
     Update
    0.66
    orso
    0.65
     quotation
    0.65
    subplot
    0.65
    Activations Density 0.075%

    No Known Activations