© 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. 10-GEMMASCOPE-RES-131K
    7. 65491
    Prev
    Next
    INDEX
    Explanations

    sequoia, sequoyah, sequined, sequela

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

    The neuron consistently activates on any token containing the letter sequence “sequ,” i.e. when it encounters that specific character string across different words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    鬟
    -2.64
    .“
    -2.47
     it
    -2.47
     The
    -2.45
     But
    -2.45
    来不及
    -2.39
     unſ
    -2.38
     ܠ
    -2.30
    ,“
    -2.22
    ”;
    -2.22
    POSITIVE LOGITS
    n
    2.64
    on
    2.56
     कोशिश
    2.41
    u
    2.39
     soir
    2.39
    of
    2.30
    ia
    2.27
    ian
    2.25
     撮っ
    2.25
     accuse
    2.22
    Activations Density 0.008%

    No Known Activations