© 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. 10517
    Prev
    Next
    INDEX
    Explanations

    list format, code separators, or command syntax

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

    The neuron strongly activates on numeric tokens—especially floating‐point numbers and other measurement/value figures.

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

    tokens that are part of punctuation marks or symbols commonly used in technical or formatted text.

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

    short grammatical function words, especially prepositions and contractions used to connect phrases.

    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
    Geplaatst
    0.81
     amorphous
    0.80
    ्य
    0.78
    ת
    0.77
     retailers
    0.73
     tributaries
    0.73
    و
    0.73
    িক
    0.72
     producers
    0.72
    ième
    0.72
    POSITIVE LOGITS
     considerados
    0.75
     pihak
    0.74
    镟
    0.71
    釘
    0.69
    ﺪ
    0.68
    parser
    0.67
     навыки
    0.66
    constructor
    0.63
    他に
    0.63
    剖
    0.63
    Activations Density 0.000%

    No Known Activations