© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 129786
    Prev
    Next
    INDEX
    Explanations

    Guanqiao, Q:, Vide,,,

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

    This neuron activates on tokens composed largely of repeated underscores (e.g. long underscore sequences used in ASCII-art headers or separators).

    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
     самая
    -2.94
     \\
    -2.58
    等は
    -2.45
     boast
    -2.41
    泚
    -2.39
     desarrollando
    -2.38
    鲻
    -2.33
    ⬈
    -2.30
    ⇹
    -2.28
    
    -2.28
    POSITIVE LOGITS
    s
    3.39
    as
    3.23
    With
    3.00
    We
    2.98
    there
    2.77
    獁
    2.67
     gelas
    2.66
    7
    2.56
    From
    2.53
    H
    2.53
    Activations Density 0.005%

    No Known Activations