© 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. 43839
    Prev
    Next
    INDEX
    Explanations

    less than or arrows

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

    The neuron fires on occurrences of the “<” symbol (and related matchers like `<~`, `~>`, `^^`, `<=`, `=>`, etc.), i.e. angle‐bracket or similar operator tokens.

    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
     to
    -2.17
     since
    -1.70
     During
    -1.66
     Agrega
    -1.55
    ?”
    -1.48
    quisar
    -1.47
     vollständ
    -1.46
    岂
    -1.45
    下さ
    -1.45
     In
    -1.45
    POSITIVE LOGITS
     baumwolle
    1.98
     venant
    1.86
    mallera
    1.79
    หรือ
    1.77
    或者
    1.77
     dekoratif
    1.73
    \
    1.73
     ٢
    1.71
     eşofman
    1.70
    '
    1.59
    Activations Density 0.029%

    No Known Activations