© 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. 38278
    Prev
    Next
    INDEX
    Explanations

    Fred Soper

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

    spaces followed by two or more consecutive spaces, which typically indicates indentation in code or formatted text.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron fires on long runs of space characters—i.e. code or text indentation/whitespace 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
     långt
    -1.45
     IntelliJ
    -1.28
    ieuses
    -1.27
    andten
    -1.26
     on
    -1.26
     V
    -1.24
     ALWAYS
    -1.22
     from
    -1.21
     conduc
    -1.20
    +'.
    -1.20
    POSITIVE LOGITS
    完成です
    1.51
     sostu
    1.49
     ludzie
    1.39
     beeind
    1.38
    することができます
    1.38
     afirmar
    1.28
     içeris
    1.27
     concluyó
    1.27
    klinik
    1.26
     borboleta
    1.25
    Activations Density 0.026%

    No Known Activations