© 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. 34-GEMMASCOPE-RES-131K
    7. 38481
    Prev
    Next
    INDEX
    Explanations

    statements of fact

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

    The neuron is highly activated by common function words—especially articles, simple linking verbs, and other frequent small words like “the,” “is,” “a,” and “just.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ▌
    -0.82
    คร
    -0.80
    део
    -0.79
     neler
    -0.75
    猛地
    -0.74
    ]+=
    -0.74
     Collegiate
    -0.73
    zew
    -0.73
    lucht
    -0.73
    }");
    -0.71
    POSITIVE LOGITS
     facts
    5.34
     fact
    4.97
    fact
    3.98
    facts
    3.95
     Facts
    3.75
    Fact
    3.73
     Fact
    3.70
    Facts
    3.64
     FACT
    3.59
    FACT
    3.34
    Activations Density 0.059%

    No Known Activations