© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 535
    Prev
    Next
    INDEX
    Explanations

    Tornado framework, breaking news, inventions

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

    The neuron strongly activates on programming‐style tokens—things like import statements, script tags, and numeric literals—i.e. bits of source‐code syntax.

    oai_token-act-pair · o4-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
     undercut
    -0.80
     kl
    -0.76
     bishops
    -0.74
    エリー
    -0.71
     pho
    -0.70
     Mauritius
    -0.70
     福岡
    -0.69
    Populations
    -0.69
    foli
    -0.69
    minecraft
    -0.68
    POSITIVE LOGITS
     Tornado
    1.03
     tornado
    0.89
    tornado
    0.87
    jine
    0.75
    Modul
    0.74
    Tornado
    0.74
     dům
    0.74
    цент
    0.73
    ongevity
    0.73
     handlers
    0.73
    Activations Density 0.008%

    No Known Activations