© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 25921
    Prev
    Next
    INDEX
    Explanations

    the first thing you do when

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

    The neuron detects phrases that introduce the first step or action in instructional or “how-to” style text (e.g. “the first thing you do,” “one of the first things we did”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     uniquement
    -1.60
     فقط
    -1.56
    recently
    -1.51
    ONLY
    -1.30
     somente
    -1.23
    currently
    -1.22
    only
    -1.20
    だけです
    -1.20
    Previously
    -1.19
    mainly
    -1.18
    POSITIVE LOGITS
     when
    2.45
     after
    2.28
     before
    1.84
     to
    1.80
     in
    1.80
     should
    1.64
     any
    1.63
     после
    1.60
     of
    1.52
     their
    1.48
    Activations Density 0.025%

    No Known Activations