© 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. 9800
    Prev
    Next
    INDEX
    Explanations

    numerical computations or specific calculations

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

    The neuron fires most strongly on adjectival words (e.g. “specific,” “public,” “private,” “their”).

    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
    Contr
    -0.77
     contu
    -0.73
    рений
    -0.73
    کارت
    -0.73
    outs
    -0.73
    бити
    -0.72
     conçus
    -0.72
     encuent
    -0.72
     deixe
    -0.71
     pengurus
    -0.71
    POSITIVE LOGITS
    راجع
    0.93
    imanapun
    0.85
    شاهد
    0.84
     clín
    0.84
     Mehta
    0.82
    capulco
    0.82
     ответствен
    0.82
    kog
    0.80
    Making
    0.79
    Putting
    0.79
    Activations Density 0.052%

    No Known Activations