© 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. 10-GEMMASCOPE-RES-131K
    7. 67933
    Prev
    Next
    INDEX
    Explanations

    Bethlehem, Palierne, Q:

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

    The neuron is detecting fragments of base64‐style or otherwise random-looking alphanumeric sequences (e.g. from encoded keys).

    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
     voor
    -2.61
    平板
    -2.19
    先天
    -2.08
    逡
    -2.06
    ?}
    -2.05
    成都
    -2.05
    ний
    -2.03
    al
    -2.02
    ziná
    -1.98
    ?)
    -1.98
    POSITIVE LOGITS
    æring
    2.08
    푣
    2.03
    总会
    1.93
    その時
    1.83
    蘼
    1.82
    越來越
    1.78
    ovács
    1.78
    *
    1.77
    渽
    1.73
    之後
    1.72
    Activations Density 0.002%

    No Known Activations