© 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. 74873
    Prev
    Next
    INDEX
    Explanations

    year-month-day ranges

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

    The neuron fires most strongly on small (mostly three‐letter) alphabetic tokens — e.g. ISO language codes, short path or file‐name pieces, domain name fragments, etc.

    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
    傚
    -1.34
    וּ
    -1.27
     frontera
    -1.23
     banyo
    -1.20
     весна
    -1.20
    ЕТ
    -1.20
    wardship
    -1.16
    itaji
    -1.16
     estekak
    -1.15
     lettres
    -1.15
    POSITIVE LOGITS
    5
    1.31
    7
    1.30
     &
    1.29
    9
    1.28
    4
    1.27
    2
    1.26
    3
    1.24
    1
    1.23
     etc
    1.21
    /
    1.16
    Activations Density 0.001%

    No Known Activations