© 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. 67476
    Prev
    Next
    INDEX
    Explanations

    specifying technical terms or proper nouns

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

    The neuron fires on occurrences of the mathematical exponent variables “p” (and its conjugate “q”) in formulaic contexts (e.g. p‐norms, subscripts, and related expressions).

    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
     groei
    -0.96
     几何
    -0.95
    guenos
    -0.95
     to
    -0.94
     SUSP
    -0.93
    💀
    -0.92
    ielded
    -0.91
    ഢ
    -0.90
     uefa
    -0.90
    よし
    -0.88
    POSITIVE LOGITS
     mêmes
    1.15
    がありますが
    1.08
    取决于
    1.05
    は
    1.05
     использование
    1.02
     mikrofon
    1.00
    <bos>
    0.99
    vedať
    0.97
     دیکھیے
    0.96
    même
    0.96
    Activations Density 0.012%

    No Known Activations