© 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. 34-GEMMASCOPE-RES-131K
    7. 78276
    Prev
    Next
    INDEX
    Explanations

    Q: followed by text

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

    This neuron fires on code‐style punctuation and symbols—especially underscores, parentheses, hash signs, etc.—i.e. it flags snippets of programming source text.

    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
     trầm
    -0.91
    getAttributes
    -0.77
     Commodore
    -0.76
     rivets
    -0.75
     enclosing
    -0.74
     pseudonym
    -0.71
    siones
    -0.69
    aderno
    -0.69
     oracle
    -0.69
     whiskey
    -0.69
    POSITIVE LOGITS
    Mice
    0.82
    Personally
    0.79
     acab
    0.78
     síntomas
    0.77
    CNN
    0.77
     mice
    0.77
     DLC
    0.75
    
    0.73
     annum
    0.73
    ibid
    0.73
    Activations Density 0.000%

    No Known Activations