© 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. 70603
    Prev
    Next
    INDEX
    Explanations

    doctrine, host, technologies, vision, controlled

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

    The neuron fires strongly on isolated, content‐bearing nouns—especially single technical or conceptual keywords (like “doctrine,” “host,” “system,” “globe,” “ignition,” etc.) that often head or label a section or topic.

    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
    ↵↵
    -1.00
    ?
    -0.95
     Также
    -0.94
     tull
    -0.91
    You
    -0.91
    😱
    -0.91
     kark
    -0.90
    istani
    -0.88
    ID
    -0.88
    。。。
    -0.87
    POSITIVE LOGITS
    鲼
    1.20
    expressing
    1.16
     Naciones
    1.13
    もありますが
    1.10
    怎么回事
    1.05
     couv
    1.05
    ewon
    1.04
     czarny
    1.03
     srebr
    1.01
     escolh
    1.01
    Activations Density 0.032%

    No Known Activations