© 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. 39938
    Prev
    Next
    INDEX
    Explanations

    status =

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

    The neuron chiefly responds to sequences of digits—that is, numeric tokens in the text.

    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
    TBA
    -0.92
    rando
    -0.91
     our
    -0.90
    restler
    -0.89
     blah
    -0.89
    liters
    -0.86
     mené
    -0.84
    緞
    -0.84
     Very
    -0.84
     or
    -0.84
    POSITIVE LOGITS
    becomes
    0.96
    remains
    0.95
     necessidade
    0.94
     furiously
    0.92
    atelli
    0.90
     フェ
    0.89
     devient
    0.88
     pozdě
    0.88
    imentel
    0.88
    ensitivity
    0.86
    Activations Density 0.020%

    No Known Activations