© 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. 391
    Prev
    Next
    INDEX
    Explanations

    code and data structures

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

    The neuron fires on identifier‐like tokens that end in numeric suffixes (e.g. “L316”, “A332”, “C292”).

    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
     the
    -2.55
     For
    -2.34
     Although
    -2.31
     because
    -2.30
     Both
    -2.22
     While
    -2.20
     after
    -2.19
     as
    -2.19
     before
    -2.16
    n
    -2.13
    POSITIVE LOGITS
    厹
    2.80
    辚
    2.67
    缡
    2.61
    ⁛
    2.50
     ſever
    2.42
    🫃
    2.41
    艄
    2.38
    氿
    2.38
    ſſen
    2.36
    ܆
    2.34
    Activations Density 0.023%

    No Known Activations