© 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. 34-GEMMASCOPE-RES-131K
    7. 94101
    Prev
    Next
    INDEX
    Explanations

    non- words

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

    This neuron fires on hyphenated or multi‐part technical compound terms (e.g. “amyloid-beta,” “non‐protein‐coding,” “non‐starch polysaccharides,” “non‐collagenous,” “non‐cardiac,” 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_34/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
    ごちそうさまでした
    -1.14
     most
    -0.98
     uncertain
    -0.94
    異なる
    -0.93
    お疲れ
    -0.93
    最も
    -0.92
    llywood
    -0.92
     gerçekten
    -0.91
     років
    -0.91
     jSON
    -0.91
    POSITIVE LOGITS
    Non
    1.20
     Non
    1.01
    non
    1.01
    cannot
    0.95
    <bos>
    0.93
    having
    0.88
     وغير
    0.88
    而是
    0.88
     Allgeme
    0.88
     Gründe
    0.85
    Activations Density 0.118%

    No Known Activations