© 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. 24-GEMMASCOPE-2-RES-16K
    7. 11705
    Prev
    Next
    INDEX
    Explanations

    language variations

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

    The neuron selectively fires on numeric or measurement‐style tokens—i.e. numbers (especially with decimal points) and related quantitative data.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    the word "of" appearing in certain generic or repetitive contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    references to quantities or amounts (numbers, doubles, deficits) within formal or instructional prose.

    oai_token-act-pair · gpt-5Triggered 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
    eter
    0.79
    ission
    0.77
    io
    0.75
    冑
    0.73
    ie
    0.72
    angular
    0.71
    תן
    0.71
    ８
    0.71
    radiation
    0.71
    esome
    0.70
    POSITIVE LOGITS
     электрон
    0.77
     п
    0.75
     multiv
    0.75
     лест
    0.73
    ленность
    0.71
     shortlist
    0.71
     therapeut
    0.69
     vipp
    0.69
     کہاں
    0.69
     seminars
    0.68
    Activations Density 0.000%

    No Known Activations