© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 127990
    Prev
    Next
    INDEX
    Explanations

    counting numbers

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

    The neuron activates on numeric tokens—especially statistical figures and decimal values—in sports and biographical data.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    のほうが
    -1.38
     zrobić
    -1.23
    ngths
    -1.21
    訫
    -1.20
     galeria
    -1.17
    átu
    -1.16
    ăzi
    -1.16
    ൎ
    -1.16
    麹
    -1.16
    过了
    -1.16
    POSITIVE LOGITS
     this
    1.49
     your
    1.45
     Seperti
    1.30
    素晴らしい
    1.28
     or
    1.28
     lavoratori
    1.26
    our
    1.25
     Spazier
    1.25
     on
    1.22
     such
    1.22
    Activations Density 0.025%

    No Known Activations