© 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. 22-GEMMASCOPE-RES-131K
    7. 26767
    Prev
    Next
    INDEX
    Explanations

    press release

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

    The neuron detects occurrences of the word “press” (often capitalized) marking press releases or press‐section headings.

    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
     
    -3.05
    [
    -2.89
    ‘
    -2.77
     ‘
    -2.75
    \
    -2.58
    (
    -2.53
    —
    -2.52
     was
    -2.50
    Y
    -2.39
    …
    -2.39
    POSITIVE LOGITS
    嗞
    2.64
     doigts
    2.53
    émoc
    2.47
     rajut
    2.45
    ଞ
    2.42
     любовь
    2.41
    鬭
    2.41
     стоковые
    2.41
     県
    2.39
    涠
    2.38
    Activations Density 0.023%

    No Known Activations