© 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. 10-GEMMASCOPE-RES-131K
    7. 105576
    Prev
    Next
    INDEX
    Explanations

    Q: technical terms or lists

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

    opening parentheses or colons that introduce technical terms, abbreviations, or structured metadata fields.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron fires on uppercase acronyms or section‐label tokens (multi‐letter capital abbreviations often followed by punctuation) that mark headings or structural labels in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    It
    -2.52
     When
    -2.50
     after
    -2.36
     or
    -2.31
    </em>
    -2.31
     For
    -2.31
     before
    -2.28
     with
    -2.23
    Not
    -2.23
     At
    -2.20
    POSITIVE LOGITS
     deliciously
    2.69
     quintessential
    2.67
     accentu
    2.63
     bloquer
    2.41
     predominantly
    2.33
    踯
    2.33
     unbelievably
    2.27
    刄
    2.27
     soooo
    2.27
     vijf
    2.27
    Activations Density 0.352%

    No Known Activations