© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 14528
    Prev
    Next
    INDEX
    Explanations

    soul and R&B artists

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

    The neuron is primarily detecting capitalized proper names and uppercase abbreviations (i.e. named entities).

    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
    着自己的
    -1.15
    怎么做
    -1.15
     juez
    -1.13
    厂商
    -1.11
    観客
    -1.08
     девочки
    -1.08
     provinces
    -1.08
     italianos
    -1.06
    代价
    -1.06
    学ぶ
    -1.06
    POSITIVE LOGITS
     afro
    1.60
     african
    1.58
     African
    1.55
    African
    1.30
    african
    1.13
     Afro
    1.13
    >("
    1.11
     caribbean
    1.09
     Baltimore
    1.08
     black
    1.07
    Activations Density 0.115%

    No Known Activations