© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 38978
    Prev
    Next
    INDEX
    Explanations

    change is good

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

    The neuron fires on subjective evaluative language—especially positive assessments like “good,” “intriguing,” “enjoying,” “benefit,” “make sense,” etc.

    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
    還要
    -0.97
    hofen
    -0.92
     ہیں۔
    -0.88
     Архивировано
    -0.88
     hort
    -0.85
    Ir
    -0.78
    }$.\\
    -0.77
    "]="
    -0.77
    '".
    -0.77
    还要
    -0.76
    POSITIVE LOGITS
     perhaps
    1.17
     maybe
    1.13
     меньше
    1.03
     actually
    1.00
    かえ
    0.95
     instead
    0.92
    perhaps
    0.89
    ͊
    0.88
     now
    0.88
    zert
    0.88
    Activations Density 0.020%

    No Known Activations