© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 60557
    Prev
    Next
    INDEX
    Explanations

    through the roof and floor

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

    This neuron responds to phrasing that signals strong intensity or extremeness (e.g. “through the roof,” “rising,” “stress,” “ringing”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    cidos
    -0.90
    IVIDUAL
    -0.89
    низу
    -0.86
    siert
    -0.84
     arterio
    -0.83
    vignon
    -0.82
    èces
    -0.81
    hnout
    -0.81
    úly
    -0.80
     Lumin
    -0.78
    POSITIVE LOGITS
     off
    3.03
     through
    1.99
    Off
    1.78
     Off
    1.73
     OFF
    1.60
    Through
    1.56
    off
    1.55
    through
    1.46
     attraverso
    1.44
     THROUGH
    1.44
    Activations Density 0.024%

    No Known Activations