© 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. 54167
    Prev
    Next
    INDEX
    Explanations

    different contexts

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

    The neuron strongly activates on words that occur at the start of new sentences (i.e. sentence-initial, capitalized or discourse-marker tokens).

    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
     on
    -1.88
     sehr
    -1.77
     at
    -1.69
     revital
    -1.66
     for
    -1.64
     epitom
    -1.62
     as
    -1.62
     in
    -1.60
     große
    -1.59
     first
    -1.58
    POSITIVE LOGITS
     imprimer
    1.98
     mauva
    1.81
     auffi
    1.80
     maniere
    1.68
     verhind
    1.67
     sooo
    1.66
     promesse
    1.66
     certamente
    1.66
     unbelievably
    1.64
     jezt
    1.64
    Activations Density 0.049%

    No Known Activations