© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 9880
    Prev
    Next
    INDEX
    Explanations

    counting heads and legs

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

    The neuron detects second‐person address (e.g. “you,” “may,” “are”)—instances where the text is speaking directly to the reader.

    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
    و
    0.86
    后期
    0.82
    ه‌ی
    0.77
    cı
    0.75
    ش
    0.75
    弼
    0.73
    ه
    0.73
     なっ
    0.73
    ی
    0.73
    ļ
    0.72
    POSITIVE LOGITS
    SHE
    0.88
    BURG
    0.87
     möchten
    0.87
    NOW
    0.82
    SOM
    0.79
    MessageNow
    0.77
    BERG
    0.77
     NOW
    0.76
    ING
    0.75
    가능
    0.75
    Activations Density 0.001%

    No Known Activations