© 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. 79525
    Prev
    Next
    INDEX
    Explanations

    Q: questions and headers

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

    The neuron strongly activates on numeric tokens—especially years and version numbers—appearing in copyright and license headers.

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

    boilerplate license/copyright headers and vendor SDK driver code in C/C++ source files (e.g., copyright years, license identifiers, and company names).

    oai_token-act-pair · gpt-5
    Triggered by @jyhe0408

    copyright notices and legal text in software source code files.

    oai_token-act-pair · claude-4-5-sonnet
    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
     crea
    -0.74
     Pickle
    -0.73
     foreigner
    -0.69
     pickle
    -0.68
     creare
    -0.68
    gly
    -0.68
     ts
    -0.68
     tambi
    -0.68
     ditt
    -0.66
    ennio
    -0.66
    POSITIVE LOGITS
     Mele
    0.93
     nalazi
    0.87
     Kine
    0.77
    MQ
    0.76
    METRO
    0.74
     MQ
    0.74
    вании
    0.73
    Fre
    0.73
     promocional
    0.72
    andung
    0.71
    Activations Density 0.032%

    No Known Activations