© 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. 31-GEMMASCOPE-2-RES-16K
    7. 1898
    Prev
    Next
    INDEX
    Explanations

    best performers and quality

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

    The neuron fires strongly on comparative and superlative degree words (e.g. more, most, less, great, best) signaling extreme or intensified descriptions.

    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_31_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
     применя
    0.65
    С
    0.62
     снаб
    0.55
    periodic
    0.55
     предусматри
    0.54
     microorganisms
    0.54
     periodically
    0.52
    с
    0.52
     устанавли
    0.51
     используют
    0.50
    POSITIVE LOGITS
     idyllic
    0.69
     wenigen
    0.65
     underwhelming
    0.64
     heutigen
    0.64
     ilk
    0.63
     größten
    0.61
     unlikely
    0.60
     besten
    0.59
     Elsewhere
    0.59
     Rookie
    0.59
    Activations Density 0.579%

    No Known Activations