© 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. 13456
    Prev
    Next
    INDEX
    Explanations

    INSTALLING PACKAGES AND CODE

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

    The neuron strongly activates on the very first word token of each new text segment (i.e. the token right after a <bos>).

    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
    м
    0.86
     отста
    0.84
    iw
    0.83
     alis
    0.81
    א
    0.80
    osw
    0.79
     сферы
    0.79
     MIA
    0.79
    рый
    0.78
     filas
    0.78
    POSITIVE LOGITS
    শৃঙ্খলা
    0.73
     gearbox
    0.73
     boots
    0.70
    मेर
    0.69
    Ձ
    0.68
     Diaries
    0.68
     Tentang
    0.67
     Players
    0.66
    cretsiz
    0.64
    আপ
    0.63
    Activations Density 0.015%

    No Known Activations