© 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. 26880
    Prev
    Next
    INDEX
    Explanations

    legal/technical document headers

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

    The neuron activates on unusually long sequences of whitespace or indentation in the text.

    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
    0
    -2.92
    各样的
    -2.67
    </b>
    -2.64
    ならば
    -2.59
    “,
    -2.53
     verminder
    -2.38
     lej
    -2.36
    あんまり
    -2.34
    these
    -2.33
     PROVID
    -2.31
    POSITIVE LOGITS
    讴
    3.16
    鐧
    2.80
    摶
    2.78
    垍
    2.66
    倶
    2.66
    瓩
    2.64
     extremist
    2.61
    谌
    2.50
     ,‬
    2.50
    觐
    2.47
    Activations Density 0.032%

    No Known Activations