© 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. 95635
    Prev
    Next
    INDEX
    Explanations

    specific word transitions

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

    This neuron detects prominent section or label tokens—the key words and punctuation that introduce or headline code snippets, questions, and technical sections.

    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
     by
    -1.44
     at
    -1.39
     in
    -1.32
     all
    -1.31
     or
    -1.27
     It
    -1.26
     There
    -1.23
    That
    -1.23
     As
    -1.21
     On
    -1.20
    POSITIVE LOGITS
     différentes
    1.55
     différents
    1.39
     inond
    1.33
    鹟
    1.33
     coté
    1.32
     récol
    1.29
     mauva
    1.27
     chauss
    1.27
     horaire
    1.27
     permett
    1.25
    Activations Density 0.050%

    No Known Activations