© 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. 34-GEMMASCOPE-RES-131K
    4. 98171
    Prev
    Next
    INDEX
    Explanations

    first or early stages

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

    This neuron fires most strongly on the core verb + object phrase—the “main clause” chunk of a sentence (e.g. “start our day,” “have what is called a ‘first look’,” “escapes the rapid pulmonic degradation”).

    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
     AQU
    -0.72
    ッセ
    -0.71
    经理
    -0.68
    VK
    -0.68
    Reprodu
    -0.67
     Española
    -0.67
    _$
    -0.66
    pausal
    -0.66
    releases
    -0.65
    ந்த
    -0.65
    POSITIVE LOGITS
     First
    3.61
    First
    3.50
    first
    3.03
     first
    2.92
     FIRST
    2.63
    FIRST
    2.61
     getFirst
    2.19
    ファースト
    2.06
    getFirst
    1.99
    第一
    1.98
    Activations Density 0.055%

    No Known Activations