© 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. 60037
    Prev
    Next
    INDEX
    Explanations

    documentation, sharing instructions, attribution, or legal terms

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

    The neuron is responding to very high‐frequency tokens (i.e. common words or identifiers that occur across documents). It effectively flags “popular” tokens by giving them large activation values.

    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
     sekaligus
    -0.82
    ísticas
    -0.80
    wnętr
    -0.79
     several
    -0.78
    ztés
    -0.77
    عداد
    -0.75
    ídas
    -0.75
     many
    -0.72
     sepanjang
    -0.71
     ALL
    -0.70
    POSITIVE LOGITS
     Wikimedia
    0.91
    vido
    0.85
    antry
    0.85
    Gallery
    0.82
     yp
    0.81
    Deel
    0.80
    URIDAD
    0.77
    ilogue
    0.76
     gallery
    0.75
    ногда
    0.75
    Activations Density 0.004%

    No Known Activations