© 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. 15230
    Prev
    Next
    INDEX
    Explanations

    "answer" or "not mentioned"

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

    The neuron strongly responds to commas in the text.

    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
    纖
    0.75
    %%
    0.73
    tek
    0.72
    ть
    0.71
     высоко
    0.71
    fer
    0.68
    melding
    0.68
     activate
    0.67
    detect
    0.67
    alarm
    0.67
    POSITIVE LOGITS
    䅞
    0.82
    渽
    0.78
     aptly
    0.75
    琊
    0.74
    ⿴
    0.72
    洨
    0.72
    وفة
    0.71
     Ud
    0.71
     संबोधित
    0.71
     复
    0.71
    Activations Density 0.000%

    No Known Activations