© 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. 74893
    Prev
    Next
    INDEX
    Explanations

    specific introductory tokens

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

    The neuron activates on numeric literals (i.e. tokens consisting of digits).

    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
    -1.66
     this
    -1.65
     at
    -1.59
     it
    -1.48
     or
    -1.44
     in
    -1.37
    為に
    -1.36
     for
    -1.35
     by
    -1.30
     will
    -1.26
    POSITIVE LOGITS
     granito
    1.36
     alp
    1.30
    ppings
    1.29
     oliveira
    1.27
     emble
    1.25
    struzione
    1.24
     錠
    1.24
     camiset
    1.24
     geograf
    1.24
    therners
    1.23
    Activations Density 0.000%

    No Known Activations