© 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. 65269
    Prev
    Next
    INDEX
    Explanations

    one or a few of

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

    The main thing this neuron does is detect numerical values (digits, numbers, and numeric measurements).

    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
     and
    -1.20
     gynnwys
    -1.08
    konos
    -1.06
    IoT
    -1.06
     SEVENTH
    -1.04
     etc
    -1.03
    Kubernetes
    -1.03
     THOUS
    -1.02
     FIFTH
    -1.00
     HUND
    -1.00
    POSITIVE LOGITS
     when
    1.44
     in
    1.25
     a
    1.23
     on
    1.17
     after
    1.09
    个月
    1.08
     during
    1.07
     shortly
    1.06
     europea
    0.96
     enfermo
    0.93
    Activations Density 0.073%

    No Known Activations