© 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. 379
    Prev
    Next
    INDEX
    Explanations

    date and time stamps

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

    The neuron strongly activates on article‐metadata tokens—things like dates, times, dateline markers (e.g. “Thursday,” “IST,” numeric date components), and image‐caption labels.

    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
     “
    -1.26
     $
    -1.12
     can
    -1.06
    \
    -1.06
     E
    -1.06
     r
    -1.05
     A
    -1.02
    š
    -0.99
     (
    -0.96
     an
    -0.96
    POSITIVE LOGITS
    ஶ
    1.28
     umě
    1.22
     chuy
    1.22
     zařízení
    1.20
     habido
    1.16
     Septiembre
    1.13
     grabar
    1.13
     hinh
    1.13
     ayudado
    1.13
     entendido
    1.12
    Activations Density 0.007%

    No Known Activations