© 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. 34-GEMMASCOPE-RES-131K
    7. 86952
    Prev
    Next
    INDEX
    Explanations

    technical jargon and code

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

    The neuron fires most strongly on code‐ or config-style metadata tokens—multi-digit version numbers, dependency or help directives, license keywords, caret (^) version specifiers, etc. In other words, it spots code/configuration header elements.

    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
     replace
    -0.74
     ден
    -0.71
     offering
    -0.69
    global
    -0.69
     Stay
    -0.69
     meta
    -0.69
    davis
    -0.69
     leps
    -0.67
     Replace
    -0.67
    stay
    -0.67
    POSITIVE LOGITS
    HSP
    0.74
    permum
    0.73
     fwd
    0.73
     jac
    0.72
     logits
    0.71
     jw
    0.70
     zodiaco
    0.70
     typescript
    0.70
     qc
    0.70
     quoi
    0.69
    Activations Density 0.087%

    No Known Activations