© 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. 33990
    Prev
    Next
    INDEX
    Explanations

    negative sentiment words

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

    the neuron detects strongly subjective or emotionally charged words (i.e. evaluative sentiment terms).

    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
     permasalahan
    -1.02
     our
    -0.92
    %;
    -0.85
     sjuk
    -0.85
    congress
    -0.85
     or
    -0.83
     earlier
    -0.83
     engraçadas
    -0.82
     tidligere
    -0.82
    
    -0.81
    POSITIVE LOGITS
    ilizar
    0.94
    Then
    0.83
    Contrast
    0.82
    ̣ng
    0.79
    deras
    0.79
     هیچ
    0.78
    нимание
    0.78
    gido
    0.76
     sitten
    0.76
    cotta
    0.75
    Activations Density 0.036%

    No Known Activations