© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 39773
    Prev
    Next
    INDEX
    Explanations

    return programming expressions

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

    The neuron is picking out the user‐defined identifiers (variable, function, and other symbol names) in code.

    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
    șit
    -0.86
    omg
    -0.85
     Goy
    -0.82
    ⁶
    -0.82
    ərb
    -0.82
    }=\
    -0.81
    iso
    -0.80
    Roboto
    -0.78
     DRAW
    -0.78
     Blüten
    -0.78
    POSITIVE LOGITS
     palsu
    0.97
     provide
    0.94
     neceffary
    0.93
     matka
    0.93
     neceff
    0.93
    richText
    0.93
     luminosa
    0.93
    ajikan
    0.92
     plancha
    0.92
     Nug
    0.92
    Activations Density 0.005%

    No Known Activations