© 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. 31-GEMMASCOPE-2-RES-16K
    7. 540
    Prev
    Next
    INDEX
    Explanations

    no single

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

    The neuron is detecting extremely common, high‐frequency function words (e.g. “is,” “there,” “a,” “also”).

    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
    ::
    0.39
     svojih
    0.37
    =
    0.37
     imong
    0.35
     ("
    0.35
     کے
    0.34
     savo
    0.34
    vist
    0.34
    ::_
    0.34
     svojoj
    0.33
    POSITIVE LOGITS
     plenty
    0.92
     no
    0.89
     nothing
    0.71
     downsides
    0.69
     lots
    0.68
     similarities
    0.68
     discrepancies
    0.68
     ample
    0.67
     возможность
    0.66
     uncertainty
    0.63
    Activations Density 0.704%

    No Known Activations