© 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. 22-GEMMASCOPE-RES-131K
    7. 66514
    Prev
    Next
    INDEX
    Explanations

    I was, I had

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

    The neuron is encoding each token’s absolute position index in the document (i.e. it “finds” the token’s sequential position number).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    شنایی
    -1.05
    ១
    -0.98
    árt
    -0.98
    ḫ
    -0.98
    一郎
    -0.97
    ány
    -0.96
     сегодняшний
    -0.96
     InputMaybe
    -0.96
    Crear
    -0.95
     characters
    -0.94
    POSITIVE LOGITS
    ilustración
    0.98
    ッキリ
    0.96
     reduzir
    0.92
     deswegen
    0.91
    eror
    0.91
     ajudá
    0.91
     autonomie
    0.90
    倘
    0.89
     какой
    0.88
     обычно
    0.87
    Activations Density 0.038%

    No Known Activations