© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128821
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly fires on isolated single‐character tokens (letters or digits) used as list or option markers (e.g. “(a)”, “(b)”, “c”, “6”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    digits and short words

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     S
    -2.11
     M
    -2.06
     K
    -1.95
     D
    -1.89
     T
    -1.88
     However
    -1.88
     A
    -1.88
     N
    -1.85
     on
    -1.83
     J
    -1.82
    POSITIVE LOGITS
    ↵↵
    2.14
    ↵
    1.90
    自分の
    1.71
     strå
    1.66
    ↵↵↵
    1.63
    继续
    1.62
     første
    1.62
     första
    1.61
    两个
    1.59
    因为
    1.55
    Activations Density 0.113%

    No Known Activations