© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 8707
    Prev
    Next
    INDEX
    Explanations

    I think

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

    The neuron detects first‐person opinion or hedging phrases (e.g. “I think,” “I believe,” expressions of personal judgment).

    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
    ли
    1.32
     어떻게
    0.96
    τητα
    0.95
    如何在
    0.91
    lı
    0.91
    如何
    0.90
    і
    0.90
    com
    0.89
     how
    0.89
    ien
    0.89
    POSITIVE LOGITS
     людьми
    0.98
     grot
    0.91
    rated
    0.90
     bijna
    0.90
     gasp
    0.90
     faceted
    0.89
     unusually
    0.88
    ف
    0.88
     faucibus
    0.88
     devemos
    0.88
    Activations Density 0.040%

    No Known Activations