© 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. 93120
    Prev
    Next
    INDEX
    Explanations

    life span development

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

    numerical years or dates in educational and technical content.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    This neuron fires on document headings and titles—i.e. section‐label tokens such as slide or module names, topic headers, and other title‐style words and numbers.

    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
     CHRISTMAS
    -1.50
     experien
    -1.38
    AIRE
    -1.38
    
    -1.32
     vives
    -1.32
     ...........
    -1.32
     Boletín
    -1.30
     immen
    -1.30
     simpel
    -1.29
     sembla
    -1.28
    POSITIVE LOGITS
     uses
    1.92
     strategies
    1.84
     various
    1.70
     use
    1.65
     many
    1.63
     choosing
    1.53
     typical
    1.52
     used
    1.52
     problems
    1.51
     different
    1.49
    Activations Density 0.016%

    No Known Activations