© 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. 31-GEMMASCOPE-2-RES-16K
    7. 9070
    Prev
    Next
    INDEX
    Explanations

    specific criteria

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

    The neuron fires most strongly on isolated letter–number combinations in decimal notation—that is, tokens involved in floating-point numerical values.

    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
     ibid
    0.76
     Mgr
    0.74
     US
    0.70
     Dept
    0.70
     Messrs
    0.70
     Entretanto
    0.69
    u
    0.66
     Natl
    0.63
     Quadrat
    0.63
     Leib
    0.62
    POSITIVE LOGITS
    מה
    0.64
    ভাবে
    0.62
    𝐬
    0.62
    .,"
    0.62
    тся
    0.61
    就像
    0.61
    вото
    0.61
    ()+"
    0.60
    軌
    0.58
    ভর
    0.58
    Activations Density 0.097%

    No Known Activations