© 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. 90589
    Prev
    Next
    INDEX
    Explanations

    hilarity or comedy

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

    The neuron fires on the defining “_Title_ is/about a … who…” lead‐in sentences of plot summaries.

    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
     sickening
    -0.93
     earlier
    -0.90
     mysteriously
    -0.89
     omin
    -0.88
    되었습니다
    -0.88
    elded
    -0.86
    </strong>
    -0.85
     again
    -0.83
    kere
    -0.82
     reportedly
    -0.81
    POSITIVE LOGITS
     hilar
    1.84
     hilarious
    1.72
     hij
    1.66
     comedy
    1.65
     comedic
    1.54
     hil
    1.48
     chaos
    1.47
    hil
    1.44
    shen
    1.37
     adventures
    1.34
    Activations Density 0.066%

    No Known Activations