© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 13032
    Prev
    Next
    INDEX
    Explanations

    when or if followed by event

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

    The neuron fires strongly on subordinating conjunctions and related words that introduce conditions or temporal clauses (e.g. “if,” “when,” “upon”).

    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
    ঊ
    -0.85
    Kepala
    -0.84
    Forrás
    -0.84
     slika
    -0.83
    溧
    -0.82
    Baş
    -0.82
    UCIÓN
    -0.81
    acağ
    -0.81
    Wyatt
    -0.81
     настолько
    -0.80
    POSITIVE LOGITS
     assuming
    0.84
    ‛
    0.81
     folgte
    0.81
     including
    0.79
    ∅
    0.78
    tt
    0.77
     MATHEMATICAL
    0.76
     RAILWAY
    0.76
    gai
    0.75
     stellte
    0.73
    Activations Density 0.049%

    No Known Activations