© 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. 1600
    Prev
    Next
    INDEX
    Explanations

    breakdown with increasing

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

    The neuron responds strongly to nouns naming large physical constructions or installations (e.g., building frames, machines, servers, manor houses, large spaces).

    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
    rapper
    0.67
    rentices
    0.66
    城市的
    0.65
    彡
    0.63
    ibly
    0.63
    ੱ
    0.63
    椑
    0.61
    대전
    0.60
    atives
    0.59
    ？
    0.59
    POSITIVE LOGITS
     Ausnahme
    1.00
     exception
    0.87
    holds
    0.83
     கூடிய
    0.82
     excepción
    0.82
     no
    0.81
     emphasis
    0.80
    drawn
    0.79
    t
    0.76
     uitz
    0.75
    Activations Density 0.168%

    No Known Activations