© 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. 25586
    Prev
    Next
    INDEX
    Explanations

    system programming keywords

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

    The neuron strongly activates on numeric tokens and constants (i.e. any token containing digits) in source code.

    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
     three
    -1.11
     all
    -1.08
     boneca
    -1.05
    Yes
    -1.04
    physis
    -1.03
     sereia
    -1.03
    しかし
    -1.02
     two
    -1.00
     Yes
    -1.00
     weiß
    -0.99
    POSITIVE LOGITS
    drück
    1.15
    bzw
    1.08
     from
    1.05
    Unters
    1.03
    ֩
    1.02
    anhänger
    1.01
    Ꮠ
    1.01
    Hersteller
    1.01
    Antwort
    1.00
    kissen
    1.00
    Activations Density 0.016%

    No Known Activations