© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 104243
    Prev
    Next
    INDEX
    Explanations

    after or once followed by an action

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

    The neuron fires on temporal or sequential discourse markers (e.g. “After,” “Once,” “Then,” “As”) that introduce steps or transitions in a procedure.

    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
     možnost
    -1.27
     Überras
    -1.18
     recientes
    -1.15
     obok
    -1.15
     funkci
    -1.13
    jedno
    -1.09
     sensaciones
    -1.09
     poprzed
    -1.09
     a
    -1.09
     páns
    -1.09
    POSITIVE LOGITS
     completed
    1.29
     then
    1.23
     finished
    1.19
     eventually
    1.17
     finally
    1.17
     done
    1.13
     eventualmente
    1.07
     final
    1.06
    meisters
    1.05
     vasca
    1.03
    Activations Density 0.030%

    No Known Activations