© 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. 70
    Prev
    Next
    INDEX
    Explanations

    code structures or identifiers

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

    The neuron responds to numeric literals (sequences of digits) in the text.

    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
     Funktionalität
    -1.16
    important
    -1.10
     convid
    -1.09
    cisive
    -1.09
     grö
    -1.06
    ringen
    -1.06
     Auswirkungen
    -1.05
    widetext
    -1.05
     Antworten
    -1.03
    kirche
    -1.02
    POSITIVE LOGITS
     or
    1.38
     with
    1.31
     from
    1.09
    operatorname
    0.93
    有时
    0.87
     '%'
    0.86
    ører
    0.85
    éges
    0.85
     Contempor
    0.85
     Affection
    0.84
    Activations Density 0.002%

    No Known Activations