© 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. 6780
    Prev
    Next
    INDEX
    Explanations

    time approximations and quantities

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

    The neuron strongly activates on tokens that introduce conditions, warnings, or formal references in instructional or legal‐style text (e.g. “possible,” “use,” “section,” “regarding,” “do”).

    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
     olhando
    0.77
     fois
    0.72
     volte
    0.71
    ל
    0.71
     شبکه
    0.70
     cambiando
    0.70
    DIA
    0.70
    い
    0.70
     utilização
    0.69
     chaîne
    0.68
    POSITIVE LOGITS
     solutes
    0.84
    ্ডিং
    0.81
    سته
    0.80
    μοσ
    0.78
    ఎస్
    0.76
    lägt
    0.75
    σον
    0.74
    สนาม
    0.74
    ክምና
    0.73
    ائیگی
    0.73
    Activations Density 0.000%

    No Known Activations