© 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. 130751
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on section‐ or heading‐style tokens—labels like “Behavior,” “Suggestions,” “Part,” “Exercise,” “Challenge,” etc., that mark document sections.

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

    describing parts or functions

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    itahu
    -0.90
     waarschijnlijk
    -0.88
     этим
    -0.87
     putern
    -0.84
    Poc
    -0.84
    Lastname
    -0.82
    ätzung
    -0.82
     nejen
    -0.81
    这么说
    -0.81
     fenomeno
    -0.81
    POSITIVE LOGITS
     involving
    1.59
     namely
    1.55
     consisting
    1.51
     wherein
    1.44
     berupa
    1.35
     whereby
    1.34
     polega
    1.30
     yaitu
    1.29
     including
    1.28
     जिसमें
    1.27
    Activations Density 0.197%

    No Known Activations