© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 102798
    Prev
    Next
    INDEX
    Explanations

    early life

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

    This neuron fires on formal, topic-label or section-heading style phrases—i.e. multi-word noun‐ or concept-introducing expressions (like “Early life,” “main focus of,” “Inflammatory bowel diseases,” “long tail,” “angry response,” “run as,” “party,” 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_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
    kým
    -1.59
     like
    -1.43
    енные
    -1.38
    皺
    -1.36
     могут
    -1.36
    ს
    -1.36
    adine
    -1.35
    :「
    -1.34
    rů
    -1.34
    .
    -1.32
    POSITIVE LOGITS
    賔
    1.52
    巨大的
    1.39
     frança
    1.36
    Origine
    1.36
    男朋友
    1.34
    </sup>
    1.33
     —
    1.33
    発信
    1.32
    独特的
    1.30
    短い
    1.30
    Activations Density 0.048%

    No Known Activations