© 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. 26250
    Prev
    Next
    INDEX
    Explanations

    strong emotional reactions

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

    The neuron detects past‐participle adjectives ending in “-ed” that express personal emotional reactions (e.g. impressed, fascinated, disappointed, intrigued, shocked).

    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
    u
    -1.76
     a
    -1.49
    er
    -1.38
     difíciles
    -1.35
     suivi
    -1.35
    aéroport
    -1.34
    e
    -1.31
     different
    -1.30
     ceea
    -1.27
     atractivos
    -1.27
    POSITIVE LOGITS
     about
    1.77
    😥
    1.68
    😠
    1.64
    😞
    1.63
    cising
    1.59
     plis
    1.56
     because
    1.55
    ܖ
    1.54
     and
    1.54
    ,‎
    1.52
    Activations Density 0.031%

    No Known Activations