© 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. 22-GEMMASCOPE-RES-131K
    7. 39699
    Prev
    Next
    INDEX
    Explanations

    indicating causation

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

    The neuron responds to common high-frequency function or linking words—i.e. stop-words and discourse markers like “the,” “is,” “by,” “because,” etc.

    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
     régiment
    -1.16
    viez
    -1.11
    LSU
    -1.04
     pesawat
    -0.98
    fgaben
    -0.97
    噥
    -0.96
    *</
    -0.95
    אין
    -0.93
    ਦੀ
    -0.92
     torr
    -0.91
    POSITIVE LOGITS
     because
    1.19
     due
    1.11
     difference
    1.00
     higher
    0.99
     некоторых
    0.99
    が違う
    0.98
     debido
    0.98
    毕竟
    0.98
     Either
    0.97
    Either
    0.96
    Activations Density 0.027%

    No Known Activations