© 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. 14555
    Prev
    Next
    INDEX
    Explanations

    common word pairings

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

    The neuron fires on numeric tokens (especially measurements, counts, and decimal numbers).

    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
     DevComponents
    -1.01
    MEM
    -0.98
     pieno
    -0.98
    grees
    -0.98
    Subviews
    -0.96
     correto
    -0.95
    agreed
    -0.95
     Pued
    -0.94
    нера
    -0.94
    ➚
    -0.93
    POSITIVE LOGITS
     and
    1.34
     These
    1.31
     for
    1.31
     these
    1.22
    These
    1.16
     or
    1.13
    Often
    1.10
     whenever
    1.09
     sometimes
    1.09
     Often
    1.07
    Activations Density 0.143%

    No Known Activations