© 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. 31-GEMMASCOPE-2-RES-16K
    7. 6135
    Prev
    Next
    INDEX
    Explanations

    business listing optimization

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

    The neuron is highly sensitive to numeric tokens or numerical expressions (e.g. actual numbers, measurements, dates).

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

    the word "and" when it connects items in formal or technical lists and sequences.

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

    connective function words—mainly conjunctions and prepositions—that link clauses or phrases and signal relations like addition, choice, purpose, or extent.

    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_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
     אחר
    0.72
     Communications
    0.69
     Reproduction
    0.69
     Controversy
    0.67
     către
    0.66
     insights
    0.66
    ົາ
    0.65
    चित
    0.64
     controversy
    0.64
     Fortune
    0.64
    POSITIVE LOGITS
     गंदगी
    0.91
    plein
    0.90
    bereich
    0.89
     nettoyer
    0.88
     бола
    0.88
    잼
    0.87
    lify
    0.86
    clonal
    0.86
    lukan
    0.84
    sede
    0.84
    Activations Density 0.000%

    No Known Activations