© 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. 10999
    Prev
    Next
    INDEX
    Explanations

    fortune cookies

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

    The neuron is keyed to the relative‐pronoun “which,” i.e. clauses beginning with “which” that introduce explanatory or defining asides.

    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
    叁
    0.84
     appellees
    0.80
    धे
    0.79
     proverb
    0.79
    душ
    0.78
    अर
    0.78
     অতিক্রম
    0.77
     Whirl
    0.77
    
    0.77
     നമഃ
    0.76
    POSITIVE LOGITS
    ك
    1.02
    ;
    0.88
    ı
    0.83
     fisik
    0.82
    이면
    0.82
    la
    0.80
    ks
    0.79
    lö
    0.77
    lardan
    0.76
    .
    0.75
    Activations Density 0.001%

    No Known Activations