© 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. 12680
    Prev
    Next
    INDEX
    Explanations

    units or states

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

    This neuron consistently picks out the central keyword or topic term of a question (the main subject the user is asking about).

    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
    ようになる
    -1.01
     whoſe
    -0.99
     geringe
    -0.98
     and
    -0.96
    tification
    -0.95
     gehörte
    -0.94
    に行った
    -0.93
     يعني
    -0.91
    decap
    -0.91
     sandalias
    -0.91
    POSITIVE LOGITS
     when
    1.22
     for
    1.16
     more
    1.12
     بیشتر
    1.11
     faster
    1.08
     most
    1.04
    ၁
    1.02
    only
    1.02
     predominant
    1.02
    すべての
    1.00
    Activations Density 0.078%

    No Known Activations