© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 65700
    Prev
    Next
    INDEX
    Explanations

    various conjunctions or following words

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

    The neuron flags unusually informative or content-heavy tokens (rare nouns, verbs, numbers, and key structural words) rather than common function words.

    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
    
    -1.14
    
    -1.14
    身边的
    -1.10
     PÅ
    -1.04
    uar
    -1.04
     –
    -1.01
     ње
    -1.01
     במ
    -1.00
    his
    -1.00
     في
    -1.00
    POSITIVE LOGITS
     which
    1.54
     it
    1.53
     different
    1.46
     różnych
    1.37
     various
    1.34
     that
    1.34
     some
    1.30
     what
    1.27
     these
    1.23
     many
    1.23
    Activations Density 0.048%

    No Known Activations