© 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. 15043
    Prev
    Next
    INDEX
    Explanations

    surprise, shock, confusion

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

    The neuron highlights words that describe a character’s physical or emotional reactions (e.g. surprise, frown, blink, cheeks, dumbfounded).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    words and phrases expressing emotional reactions of shock, surprise, or being taken aback.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    descriptions of strong emotional reactions—especially surprise, shock, or amazement—often conveyed through facial expressions and body language.

    oai_token-act-pair · gpt-5Triggered 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
    今日も
    0.81
     tedious
    0.63
     spacious
    0.62
     keypad
    0.61
     workday
    0.59
     evening
    0.59
     evenings
    0.59
     reminders
    0.59
    ourage
    0.57
     reminder
    0.57
    POSITIVE LOGITS
     shocked
    1.01
    诧
    0.94
     astonished
    0.94
    惊讶
    0.93
     incred
    0.92
    震惊
    0.92
     astonishment
    0.91
     stunned
    0.91
    😲
    0.89
     Shock
    0.88
    Activations Density 0.047%

    No Known Activations