© 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. 12834
    Prev
    Next
    INDEX
    Explanations

    as is

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

    The neuron fires on high‐frequency English function words (copulas, pronouns, modals) and past‐tense or past‐participle verbs.

    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
     trám
    -0.85
     davvero
    -0.83
     DOES
    -0.77
    ||}
    -0.77
    assertNot
    -0.76
     time
    -0.75
    随后
    -0.75
     tiempo
    -0.74
    chini
    -0.73
    果然
    -0.73
    POSITIVE LOGITS
     without
    1.32
    そのまま
    1.21
    without
    1.12
    のまま
    1.09
     its
    1.01
     unchanged
    0.95
     unaltered
    0.94
    まま
    0.91
     ללא
    0.90
     unmodified
    0.90
    Activations Density 0.064%

    No Known Activations