© 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. 45434
    Prev
    Next
    INDEX
    Explanations

    legal rights and publication notices

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

    The neuron is activating most strongly on relatively uncommon, domain‐specific “jargon” words (i.e. technical or specialized vocabulary).

    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
    貧
    -0.69
    หร
    -0.68
    }+$
    -0.66
     wickedness
    -0.65
     flank
    -0.63
    Idem
    -0.63
     thảm
    -0.62
     yah
    -0.62
    good
    -0.62
    Uploader
    -0.61
    POSITIVE LOGITS
     cirugía
    0.82
    orea
    0.70
    ict
    0.68
    Genome
    0.67
     Genome
    0.66
     Clifton
    0.66
    bilidad
    0.65
    rait
    0.65
     نک
    0.65
    ં
    0.63
    Activations Density 0.070%

    No Known Activations