© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129625
    Prev
    Next
    INDEX
    Explanations

    The neuron primarily detects four‐digit numeric tokens that represent years (e.g. copyright date strings).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    US patents and copyrights

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     when
    -1.34
    yorum
    -1.22
     because
    -1.22
     after
    -1.20
    نسية
    -1.17
    饅
    -1.16
     until
    -1.14
     guitare
    -1.10
     for
    -1.05
     its
    -1.05
    POSITIVE LOGITS
     trám
    1.10
    ure
    1.10
    lier
    1.05
    ics
    1.04
     U
    1.04
    ‐
    1.02
    ustes
    1.01
    ins
    1.00
    na
    0.98
    des
    0.98
    Activations Density 0.002%

    No Known Activations