© 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. 22-GEMMASCOPE-RES-131K
    7. 13100
    Prev
    Next
    INDEX
    Explanations

    layer background

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

    The neuron detects the single most salient content word (often a central noun or topic keyword) in headings or titles.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     their
    -1.29
    chemise
    -1.23
     either
    -1.16
    pantalon
    -1.13
    armani
    -1.12
     both
    -1.10
    levis
    -1.09
     ripet
    -1.08
     angenehm
    -1.07
    monter
    -1.06
    POSITIVE LOGITS
     المياه
    1.13
     is
    1.11
     ARISING
    1.06
    韮
    1.04
    ":
    1.04
    Descriere
    1.02
    will
    1.02
    does
    1.00
     defi
    0.99
    ten
    0.98
    Activations Density 0.000%

    No Known Activations