© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 11806
    Prev
    Next
    INDEX
    Explanations

    **Broader / Lou / easy / intentional**

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

    The neuron strongly activates on words that occur at the very start of a sentence (i.e. sentence-initial tokens).

    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
    이라
    0.88
    com
    0.86
    čin
    0.82
     tiver
    0.82
     gesagt
    0.80
     될
    0.79
    事項
    0.79
    되는
    0.78
     konto
    0.78
     zorgt
    0.77
    POSITIVE LOGITS
     redundancies
    0.70
     figures
    0.69
     autoimmunity
    0.68
    终
    0.68
     ulcers
    0.68
    汽车
    0.67
     cultured
    0.67
     vents
    0.66
     toxins
    0.66
     extrac
    0.66
    Activations Density 0.000%

    No Known Activations