© 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. 60680
    Prev
    Next
    INDEX
    Explanations

    total followed by a specific noun

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

    The neuron is detecting occurrences of the word “total” (often in headings or stats) and the associated numeric totals.

    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
     antigos
    -0.81
     troubles
    -0.81
     బ
    -0.76
     повы
    -0.75
     Defic
    -0.75
    バ
    -0.75
     lê
    -0.73
     ró
    -0.73
     meditación
    -0.71
    zung
    -0.71
    POSITIVE LOGITS
     total
    1.59
    total
    1.59
     totale
    1.45
     complete
    1.45
    itarianism
    1.41
     TOTAL
    1.40
     strangers
    1.36
     Total
    1.36
    TOTAL
    1.31
     utter
    1.29
    Activations Density 0.027%

    No Known Activations