© 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. 13248
    Prev
    Next
    INDEX
    Explanations

    codes that return values

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

    The neuron fires on error-handling guards—particularly the “if (error) return …” or similar early-exit checks 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
     Sosial
    -1.11
     décret
    -0.98
     meraih
    -0.97
     assurent
    -0.97
    注意到
    -0.94
    RDD
    -0.93
    fört
    -0.93
    湏
    -0.92
    ——”
    -0.92
     memutus
    -0.92
    POSITIVE LOGITS
    return
    1.43
     return
    1.16
     would
    1.09
     to
    1.04
     then
    1.02
     work
    1.01
     but
    0.99
     will
    0.98
     only
    0.97
     trả
    0.97
    Activations Density 0.002%

    No Known Activations