© 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. 84158
    Prev
    Next
    INDEX
    Explanations

    cold followed by specific nouns

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

    This neuron activates on occurrences of the word “cold” (including variants like “cold-smoked,” “cold-trap,” “coldly,” 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
    迨
    -2.67
    '
    -2.52
    </h3>
    -2.47
     هناك
    -2.25
    :
    -2.22
     использова
    -2.16
    垍
    -1.98
    跄
    -1.93
     heisst
    -1.92
    [
    -1.92
    POSITIVE LOGITS
     یا
    2.56
    "]="
    2.47
    鼷
    2.39
    駡
    2.38
     или
    2.34
    嘬
    2.11
    ⽔
    2.11
    遹
    2.11
     memasak
    2.09
    锘
    2.08
    Activations Density 0.006%

    No Known Activations