© 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. 15997
    Prev
    Next
    INDEX
    Explanations

    vanguard

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

    This neuron fires strongly on first-person pronouns and immediate sentence-initial narrative markers (e.g. “I,” “It,” “I’m,” “I’ll”) indicating speaker-referential statements.

    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
     первой
    0.94
     FGFR
    0.93
    ковая
    0.89
     ज्ञात
    0.85
    ర్య
    0.84
    ﻅ
    0.84
     首先
    0.84
     eurozone
    0.84
     Felipe
    0.84
     ASME
    0.84
    POSITIVE LOGITS
    beau
    0.91
    ↴
    0.83
    handling
    0.78
    dynamics
    0.75
    pros
    0.75
    putting
    0.72
    sensing
    0.72
    dan
    0.71
    fors
    0.71
    tuned
    0.71
    Activations Density 0.000%

    No Known Activations