© 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. 39684
    Prev
    Next
    INDEX
    Explanations

    API imports and identifiers

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

    The neuron fires on programming‐language identifiers—especially imported symbols, class or type names, and constants (often in ALL_CAPS or camel-case with underscores).

    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
     still
    -1.01
     is
    -0.98
    /
    -0.94
     on
    -0.93
     time
    -0.92
     does
    -0.90
     usually
    -0.89
     has
    -0.85
     whatever
    -0.85
     changed
    -0.84
    POSITIVE LOGITS
    1.16
    preise
    1.09
    expiresIn
    1.05
     ndani
    1.03
     cuartos
    1.02
    便捷
    1.01
    álogos
    1.01
    1.00
    everyone
    0.99
    简洁
    0.97
    Activations Density 0.014%

    No Known Activations