© 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. 6093
    Prev
    Next
    INDEX
    Explanations

    step by step

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

    The neuron responds to occurrences of the word “step” (and its variants like “stepping” or “step‐by‐step”).

    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
     Oaks
    0.64
     SAC
    0.61
    SAC
    0.61
     NOC
    0.60
     BSC
    0.59
     sự
    0.56
     티
    0.56
     Hid
    0.55
     διαδικ
    0.55
     Mains
    0.55
    POSITIVE LOGITS
     stepping
    1.09
     stepped
    1.03
     step
    1.01
     Step
    0.99
    estep
    0.98
    stepping
    0.96
     stepfather
    0.95
    밟
    0.95
    Step
    0.93
    步
    0.92
    Activations Density 0.135%

    No Known Activations