© 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. 45274
    Prev
    Next
    INDEX
    Explanations

    analysts, experts, observers

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

    The neuron is a sentence‐boundary/connector detector—it lights up strongly on punctuation (periods, commas) and sentence-leading or clause-linking words (e.g. But, as, shot with).

    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
    äldrar
    -1.03
    });*/
    -0.94
    滏
    -0.91
    nout
    -0.90
    žd
    -0.90
    σου
    -0.89
     génér
    -0.89
     الإنسان
    -0.89
    bücher
    -0.88
    sells
    -0.88
    POSITIVE LOGITS
     experts
    4.84
     analysts
    4.34
     observers
    3.78
    experts
    3.41
     Experts
    3.20
    Experts
    3.19
     pund
    3.13
    Analysts
    3.11
     expert
    3.05
     analyst
    3.03
    Activations Density 0.120%

    No Known Activations