© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 3618
    Prev
    Next
    INDEX
    Explanations

    influences and manipulations

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

    The neuron detects high-level action or policy verbs (e.g. boycott, subsidize, revitalize, orchestrate) that call for or describe organized interventions.

    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
    គឺ
    0.67
    谛
    0.67
    ُر
    0.63
    !\
    0.61
    萨
    0.59
     लेकर
    0.58
    คือ
    0.57
    )
    0.57
    =
    0.57
    oggler
    0.57
    POSITIVE LOGITS
    ing
    1.00
    ة
    0.89
    ic
    0.76
     பகுதிகளில்
    0.73
    ités
    0.71
    ு
    0.69
    িং
    0.67
     کردن
    0.67
     Ін
    0.66
    ução
    0.65
    Activations Density 0.000%

    No Known Activations