© 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. 7850
    Prev
    Next
    INDEX
    Explanations

    winning and losing outcomes

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

    The neuron fires strongly on mentions of wins, losses, and other performance or achievement statistics (e.g. “won,” “win,” “lose,” often with numeric counts).

    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
    פּ
    1.28
     פּ
    1.26
    א
    1.22
    ی
    1.21
    ture
    1.20
     servings
    1.13
     getDateTime
    1.09
    то
    1.09
    ה
    1.06
    שׁ
    1.05
    POSITIVE LOGITS
    ani
    1.13
    ches
    1.11
    🏆
    1.06
     Eqs
    0.99
     convincingly
    0.98
     Tiles
    0.96
    ör
    0.95
     Elys
    0.95
    av
    0.93
    ios
    0.93
    Activations Density 0.224%

    No Known Activations