© 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. 59986
    Prev
    Next
    INDEX
    Explanations

    hot hot work hot linking

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

    The neuron responds strongly to very high‐frequency tokens—common words (especially “the,” “I,” “to,” etc.) and frequent small tokens like digits—i.e. it flags broadly frequent “stop” words and numbers.

    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
    澈
    -0.82
     Hakim
    -0.75
     Classifier
    -0.74
    lown
    -0.74
    Kuri
    -0.73
     प्य
    -0.71
    ाउन
    -0.70
    ovka
    -0.70
    ХА
    -0.70
    itt
    -0.70
    POSITIVE LOGITS
     hot
    4.78
    hot
    3.91
     Hot
    3.86
    Hot
    3.83
     HOT
    2.75
    HOT
    2.75
    ホット
    2.61
    热
    2.41
     горя
    2.41
     hottest
    2.16
    Activations Density 0.077%

    No Known Activations