© 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. 130597
    Prev
    Next
    INDEX
    Explanations

    The neuron detects long delimiter lines or formatting separators composed of repeated punctuation characters (e.g. sequences of “=”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    scientific and medical introductions

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

    horizontal dividing lines and section breaks in academic/scientific documents.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi
    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
    ကိုးကား
    -0.90
     autre
    -0.82
     lasci
    -0.81
    mination
    -0.81
    midos
    -0.80
     Defining
    -0.80
     IndexError
    -0.80
    ठ
    -0.79
     Vaccination
    -0.79
     denne
    -0.79
    POSITIVE LOGITS
    ólicos
    0.91
    ↵↵
    0.91
     ambayo
    0.89
     negó
    0.88
    breos
    0.86
    PHONY
    0.85
     erzä
    0.84
    CCIÓN
    0.84
    staben
    0.82
    sibilità
    0.81
    Activations Density 0.124%

    No Known Activations