© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 70798
    Prev
    Next
    INDEX
    Explanations

    epitome, epitope, epitaph

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

    The neuron detects words related to "epitope" or forms of the word like "epitaxial" or "epitomises", which are technical terms often used in scientific contexts.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi
    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
    摀
    -2.42
    !"
    -2.39
     $
    -2.30
    謌
    -2.19
    扨
    -2.17
     were
    -2.14
     arid
    -2.11
     suitably
    -2.09
    我都
    -2.08
     Betrach
    -2.06
    POSITIVE LOGITS
    [
    2.31
    e
    2.25
    '
    2.22
     jewellery
    2.16
    During
    2.13
     نحن
    2.05
    何故
    2.02
    ↵
    2.00
     пти
    1.98
     we
    1.97
    Activations Density 0.012%

    No Known Activations