© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 28491
    Prev
    Next
    INDEX
    Explanations

    hexadecimal or values

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

    The neuron strongly responds to code identifiers used as keys or attribute names (e.g. “url,” “attr,” “name,” “Mode”) in JSON, XML, or templating/markup contexts.

    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
    illing
    -0.71
    したい
    -0.68
     Synth
    -0.68
    限界
    -0.68
    </
    -0.68
     trying
    -0.68
    nici
    -0.68
    ”.
    -0.68
     try
    -0.66
    verages
    -0.66
    POSITIVE LOGITS
    housse
    0.90
    andoah
    0.89
    °•
    0.88
    xdb
    0.85
    ecimento
    0.84
    atterns
    0.82
    outchouc
    0.82
    Ň
    0.82
    ErrUnexpectedEOF
    0.82
     asyncio
    0.81
    Activations Density 0.011%

    No Known Activations