© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 113276
    Prev
    Next
    INDEX
    Explanations

    turn and directions

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

    The neuron detects instances of the word “turn” (and closely related uses like “twist”) 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_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
    {
    -1.25
    WHY
    -1.20
    Efter
    -1.13
    ↵↵↵↵↵↵↵↵↵↵↵↵
    -1.08
    Sortie
    -1.07
    ቸ
    -1.06
     Quận
    -1.05
     crescer
    -1.05
     позволяют
    -1.04
    Cie
    -1.03
    POSITIVE LOGITS
     new
    1.48
     entweder
    1.39
     from
    1.38
     +
    1.35
     either
    1.25
     around
    1.25
     richting
    1.25
    Menurutnya
    1.24
     turned
    1.22
    要么
    1.21
    Activations Density 0.029%

    No Known Activations