© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 37877
    Prev
    Next
    INDEX
    Explanations

    prior art descriptions

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

    The neuron primarily detects tokens that contain numerals or are numeric references (e.g. citation numbers, percentages, model numbers, or other digit‐based expressions).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    條
    -0.74
    σεων
    -0.71
    CHECK
    -0.69
     without
    -0.66
     Leich
    -0.66
    晗
    -0.65
    Magn
    -0.65
     from
    -0.65
     oficinas
    -0.64
     during
    -0.64
    POSITIVE LOGITS
    ところです
    0.78
     cohort
    0.76
    Donnerstag
    0.74
     bestehende
    0.73
     fracaso
    0.73
     bodice
    0.73
    gemene
    0.73
    一気に
    0.72
    人たち
    0.71
    baran
    0.70
    Activations Density 0.053%

    No Known Activations