© 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. 15194
    Prev
    Next
    INDEX
    Explanations

    other people or groups

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

    The neuron fires on words and punctuation associated with imposing financial or policy burdens—terms like “forcing,” “taxpayers,” “bail out,” “union” and heavy commas that signal those cost-related clauses.

    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
     ägg
    -0.81
    なくなって
    -0.79
     låg
    -0.77
     Gradient
    -0.77
     élect
    -0.76
     vinyle
    -0.75
    jder
    -0.75
     electing
    -0.74
     yeniden
    -0.73
    Laurence
    -0.73
    POSITIVE LOGITS
     other
    1.41
    其他
    1.16
    autres
    1.09
     safety
    1.04
     others
    1.04
     fellow
    1.03
    otros
    1.01
     anderer
    0.98
     других
    0.95
    อื่น
    0.91
    Activations Density 0.047%

    No Known Activations