© 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. 60212
    Prev
    Next
    INDEX
    Explanations

    to or too

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

    The neuron fires on the single most semantically “important” content word in each text span, effectively picking out the keyword that best characterizes that snippet.

    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
     همچنین
    -0.91
    ――――
    -0.88
     functionalities
    -0.85
    ―――
    -0.80
    Bem
    -0.79
     rythme
    -0.78
    　　-
    -0.78
     consequent
    -0.77
     copa
    -0.75
     crease
    -0.75
    POSITIVE LOGITS
     too
    3.17
    too
    1.84
    Too
    1.77
     Too
    1.62
    到的
    1.27
     слишком
    1.26
     demasiado
    1.26
    太
    1.25
    TOO
    1.23
     terlalu
    1.18
    Activations Density 0.038%

    No Known Activations