© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 9902
    Prev
    Next
    INDEX
    Explanations

    advanced civilizations or artificial intelligence

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

    The neuron consistently fires on the final word just before a punctuation mark (i.e. the clause- or sentence-ending token).

    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_24_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
     Хи
    0.86
    ungkin
    0.82
    šit
    0.80
    šanu
    0.78
     Органи
    0.78
    سين
    0.77
    
    0.77
    㨁
    0.76
    蛜
    0.76
     Chứng
    0.74
    POSITIVE LOGITS
    +
    0.94
    -
    0.92
     from
    0.86
     by
    0.83
     for
    0.81
     vs
    0.80
     of
    0.80
     or
    0.77
     version
    0.76
     his
    0.73
    Activations Density 0.000%

    No Known Activations