© 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. 10-GEMMASCOPE-RES-131K
    7. 71468
    Prev
    Next
    INDEX
    Explanations

    laughing and smiling

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

    The neuron fires on degree‐or extent‐marking modifiers (e.g. “all the way,” “little,” “entire,” “right into,” “through the whole”) that intensify or qualify an action.

    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
     to
    -2.14
     so
    -1.61
     This
    -1.48
     always
    -1.48
     do
    -1.44
    this
    -1.42
     Those
    -1.41
     not
    -1.34
     ”
    -1.30
    This
    -1.27
    POSITIVE LOGITS
    埽
    1.49
    --"
    1.45
    短暂
    1.42
    亹
    1.42
    饜
    1.38
    雱
    1.38
    ейчас
    1.35
     APPARATUS
    1.34
    泐
    1.34
    伴随着
    1.32
    Activations Density 0.031%

    No Known Activations