© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 1401
    Prev
    Next
    INDEX
    Explanations

    according to

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

    This neuron detects journalistic source attributions (e.g. “according to …” or “said …”) often followed by a numeric citation.

    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
    ו
    0.80
    و
    0.71
    =>
    0.68
     conduce
    0.66
    গরে
    0.65
    ହି
    0.62
    :
    0.61
    ুর
    0.60
    াতি
    0.60
     hổ
    0.60
    POSITIVE LOGITS
     словам
    0.92
    Reuters
    0.91
    Officials
    0.89
     Reuters
    0.82
    ключа
    0.82
     údaje
    0.81
    数据显示
    0.79
     estimates
    0.78
     CNBC
    0.78
     officials
    0.77
    Activations Density 0.013%

    No Known Activations