© 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. 847
    Prev
    Next
    INDEX
    Explanations

    might followed by verb

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

    The neuron is keyed to speculative or hedging cues—words like “might,” “maybe,” and similar uncertainty markers.

    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
     وعند
    0.80
     وعلى
    0.74
     diễn
    0.66
    ਟ
    0.66
     dương
    0.61
     DQN
    0.61
    դ
    0.61
     Validity
    0.60
     تقريبا
    0.60
    <unused2121>
    0.60
    POSITIVE LOGITS
    t
    0.84
    might
    0.75
     might
    0.74
    碍
    0.73
    ielleicht
    0.72
    mente
    0.71
    Might
    0.70
     einiger
    0.68
    未必
    0.68
    かもしれませんが
    0.68
    Activations Density 0.166%

    No Known Activations