© 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. 9049
    Prev
    Next
    INDEX
    Explanations

    memory, speculation

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

    The neuron strongly responds to descriptive adjectives and adverbs that qualify actions or states (e.g. “blissfully,” “stressed,” “pleasant,” “uninvolved”).

    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
     increíble
    0.93
    ोर
    0.82
     ફક્ત
    0.79
    🌎
    0.79
    𝑎
    0.77
     advancements
    0.77
     incroyable
    0.77
     incredible
    0.75
     incrível
    0.73
    𝑘
    0.72
    POSITIVE LOGITS
     apologised
    0.95
     spiced
    0.90
     dodgy
    0.88
    らしく
    0.83
     (?)
    0.82
     (?
    0.81
     signalled
    0.79
     由于
    0.78
    avourite
    0.77
    いますが
    0.76
    Activations Density 0.436%

    No Known Activations