© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 68215
    Prev
    Next
    INDEX
    Explanations

    questions, file paths, names, code

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

    The neuron fires on tokens containing digits or punctuation‐rich technical identifiers (e.g. version numbers, decimal coordinates, epitope or patch IDs, URLs).

    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
     ß
    -1.03
     certamente
    -1.02
     squadra
    -0.96
     -'
    -0.96
     promover
    -0.96
     normalerweise
    -0.94
     -"
    -0.94
     radeon
    -0.94
    ようで
    -0.94
     JAMES
    -0.94
    POSITIVE LOGITS
     some
    1.40
     various
    1.38
     even
    1.30
     without
    1.30
     gối
    1.27
     that
    1.25
     several
    1.23
     with
    1.17
     numerous
    1.16
     ЛИ
    1.16
    Activations Density 0.000%

    No Known Activations