© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 84458
    Prev
    Next
    INDEX
    Explanations

    first action or beginning

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

    The neuron detects imperative action words (step‐by‐step instruction verbs) in procedural or how‐to text.

    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
     those
    -1.51
     and
    -1.47
    那些
    -1.33
     other
    -1.19
    these
    -1.07
     also
    -1.04
     more
    -1.02
    innerHTML
    -1.00
    ulism
    -0.98
     Hauses
    -0.98
    POSITIVE LOGITS
     défauts
    1.47
     zunächst
    1.25
    まずは
    1.25
     introductory
    1.17
     primeiros
    1.15
     acondicion
    1.13
     first
    1.13
     eerst
    1.12
     délais
    1.12
     start
    1.11
    Activations Density 0.056%

    No Known Activations