© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 15912
    Prev
    Next
    INDEX
    Explanations

    out of place, bounds, control, memory

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

    The neuron detects instances of the idiomatic construction “out of” (as in “out of sight,” “out of mind,” “out of scope,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    َل
    0.74
    }_{
    0.64
    Lng
    0.64
    lv
    0.63
    َع
    0.63
    yng
    0.62
    البه
    0.62
    َب
    0.61
     ngừng
    0.61
    yz
    0.61
    POSITIVE LOGITS
     respect
    0.79
     bounds
    0.73
     necessity
    0.71
     reach
    0.71
     context
    0.70
     norm
    0.69
     hindsight
    0.68
     Bounds
    0.68
     sync
    0.67
     proportion
    0.65
    Activations Density 0.093%

    No Known Activations