© 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. 15062
    Prev
    Next
    INDEX
    Explanations

    more than / than some

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

    The neuron is highly sensitive to words that occur right at the start of a new segment or sentence (i.e. the very first content token after a break).

    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
    𝒚
    0.89
    é
    0.74
    WISE
    0.73
    spielen
    0.73
    en
    0.72
    ం
    0.71
    elijke
    0.71
     обеспечение
    0.71
    placed
    0.69
    elb
    0.68
    POSITIVE LOGITS
    新款
    0.78
     workmanship
    0.73
    ड्डी
    0.73
     noun
    0.70
     ১২
    0.70
     carnage
    0.69
    Buf
    0.68
    粼
    0.68
    Sev
    0.68
     manfaat
    0.67
    Activations Density 0.000%

    No Known Activations