© 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. 10-GEMMASCOPE-RES-131K
    7. 104347
    Prev
    Next
    INDEX
    Explanations

    transmembrane protein

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

    scientific or technical terms that end in specific syllables like "ine," "ment," "ism," "oid," or "ad."

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron fires strongly on mid‐word pieces of longer, multi­syllable or technical terms (i.e. subword stems of uncommon or formal words).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     sonhos
    -1.70
    呖
    -1.64
    見る
    -1.64
     perdidos
    -1.63
    别的
    -1.63
     otras
    -1.60
     preparados
    -1.60
     subido
    -1.60
     detectar
    -1.59
     fatos
    -1.58
    POSITIVE LOGITS
    '
    2.23
     whom
    1.70
     tlie
    1.65
     we
    1.59
    ECAUSE
    1.54
     ihe
    1.50
     :}
    1.49
    鬭
    1.48
    を用
    1.46
    ୈ
    1.45
    Activations Density 0.032%

    No Known Activations