© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 34624
    Prev
    Next
    INDEX
    Explanations

    cargo cults, all together, "Kyu nahin?", "Paisa"

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

    The neuron fires on tokens that are part of structural or formatting markup—things like titles, URLs, headings, proper‐name fragments, hyphens/underscores, and other unusual punctuation sequences.

    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
    ocardial
    -1.12
    私たち
    -1.10
    Selain
    -1.08
     agréable
    -1.07
    Hazardous
    -1.05
     ABBREVIATIONS
    -1.04
    Couldn
    -1.02
    nteral
    -1.02
    uyla
    -1.00
    Heading
    -1.00
    POSITIVE LOGITS
     many
    1.52
     very
    1.48
     strong
    1.44
     big
    1.41
     good
    1.39
     no
    1.29
     like
    1.24
     now
    1.14
     give
    1.13
     when
    1.12
    Activations Density 0.015%

    No Known Activations