© 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. 72233
    Prev
    Next
    INDEX
    Explanations

    starts with "so"

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

    This neuron fires on the opening clause of a “So …” sentence—i.e. the first few words of a new paragraph that begin with “So” (e.g. “So I have really …,” “So while …,” etc.).

    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
     its
    -1.30
     ţ
    -1.28
     pourtant
    -1.05
     său
    -1.03
    ㊣
    -1.01
     coarse
    -0.98
     puertos
    -0.98
     stronę
    -0.97
    本章未完
    -0.96
    めました
    -0.96
    POSITIVE LOGITS
     what
    1.60
     basically
    1.43
    basically
    1.30
    Basically
    1.30
     now
    1.27
     obviously
    1.12
     finally
    1.02
     tersangka
    0.99
     Basically
    0.98
    Essentially
    0.98
    Activations Density 0.061%

    No Known Activations