© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 39700
    Prev
    Next
    INDEX
    Explanations

    generally speaking, rule

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

    This neuron detects hedge phrases that signal a generalization (e.g. “generally speaking” or “as a general rule”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     –
    -2.13
     and
    -1.70
     For
    -1.68
    ↵↵
    -1.67
    💞
    -1.57
     to
    -1.56
    ???:
    -1.55
     on
    -1.43
    代わりに
    -1.43
    northeast
    -1.41
    POSITIVE LOGITS
     allarg
    1.92
     gese
    1.88
    j
    1.84
    3
    1.82
     Vielen
    1.81
    of
    1.80
    However
    1.79
    很少
    1.77
     raffredd
    1.74
    c
    1.72
    Activations Density 0.001%

    No Known Activations