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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 48835
    Prev
    Next
    INDEX
    Explanations

    legal and license text

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

    The neuron strongly activates on numeric literals (e.g. version numbers, sizes, checksums) appearing in the 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_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
    眺望
    -0.76
    Alguna
    -0.76
     utan
    -0.76
    LoggedIn
    -0.72
     found
    -0.70
    øy
    -0.69
     तीन
    -0.69
    “……”
    -0.69
     viciss
    -0.69
     unrelenting
    -0.68
    POSITIVE LOGITS
     правительства
    0.87
     /**
    0.87
    
    0.83
     KAN
    0.82
    zzar
    0.82
    齑
    0.81
     ministres
    0.81
    },"
    0.81
    küche
    0.81
    NgModule
    0.80
    Activations Density 0.002%

    No Known Activations