© 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. 75623
    Prev
    Next
    INDEX
    Explanations

    Plaintiff and followed by and

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

    The neuron responds to uppercase‐heavy tokens—things like code keywords and macros (e.g. DEFINE, AMD, MEDIA) and legal/role labels in all caps (e.g. PLAINTIFF, DEFENDANT, RESPONDENT).

    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
     &
    -2.50
     \&
    -1.34
     &[
    -1.34
    (&
    -1.31
     "&
    -1.26
    ,&
    -1.24
    '&
    -1.21
     &,
    -1.21
     (&
    -1.18
    &
    -1.18
    POSITIVE LOGITS
     and
    1.48
    和
    1.19
    and
    1.19
    她和
    1.09
     And
    1.09
    And
    1.02
    我和
    1.00
     এবং
    0.96
     AND
    0.94
    และ
    0.92
    Activations Density 0.312%

    No Known Activations