© 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. 22-GEMMASCOPE-RES-131K
    7. 13927
    Prev
    Next
    INDEX
    Explanations

    comment separators

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

    The neuron fires on long runs of repeated punctuation (e.g. lines of dashes or asterisks), i.e. horizontal separator blocks.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    u
    -2.70
    o
    -2.23
    e
    -2.11
    D
    -1.99
    s
    -1.95
     GetComponent
    -1.90
    N
    -1.90
    广西
    -1.89
    O
    -1.85
    y
    -1.76
    POSITIVE LOGITS
    ↵
    3.13
    .
    2.23
     (“
    2.22
    瘜
    2.16
    髕
    2.06
    髭
    1.97
     และ
    1.96
    ftrag
    1.95
     WHICH
    1.94
     abriu
    1.93
    Activations Density 0.046%

    No Known Activations