© 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. 36114
    Prev
    Next
    INDEX
    Explanations

    will likely happen

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

    The neuron fires most strongly on semantically salient, high–information‐content words (i.e. uncommon or content-bearing tokens).

    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
    elsif
    -1.14
    providedIn
    -1.05
    ियो
    -0.98
     lavable
    -0.93
    ICIAN
    -0.93
    рі
    -0.92
    меты
    -0.91
     nød
    -0.90
     deepcopy
    -0.90
     рік
    -0.87
    POSITIVE LOGITS
     would
    1.38
     will
    1.34
     then
    1.10
     certainly
    1.09
    eschön
    1.04
     instead
    1.02
     definitely
    1.02
    certainly
    0.99
     selbst
    0.98
    首先
    0.96
    Activations Density 0.002%

    No Known Activations