© 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. 12520
    Prev
    Next
    INDEX
    Explanations

    good followed by range, taste, code, weather, practices

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

    The neuron fires strongly on positive evaluative adjectives conveying praise or approval (e.g. “good,” “great,” “better,” “fantastic,” “ideal”).

    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
    ১০
    1.39
    j
    1.16
     DiCaprio
    1.12
     Şu
    1.04
    ่
    1.01
    lu
    1.00
    loten
    0.99
    də
    0.99
     diseases
    0.98
    ০
    0.96
    POSITIVE LOGITS
    상을
    1.14
     खासा
    1.09
    ست
    1.06
     качества
    0.98
    ắn
    0.96
    삭
    0.96
    ية
    0.90
     suited
    0.90
     fortune
    0.90
    นี่
    0.89
    Activations Density 0.844%

    No Known Activations