© 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. 93324
    Prev
    Next
    INDEX
    Explanations

    class and enum declarations

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

    The neuron fires strongest on tokens that name or define code entities (e.g. class, module, package, or namespace identifiers).

    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
     first
    -1.28
     this
    -1.09
     chose
    -1.09
     people
    -1.09
     time
    -1.07
     aforesaid
    -1.05
    It
    -1.02
     came
    -1.01
     business
    -1.00
     serves
    -1.00
    POSITIVE LOGITS
    cannot
    1.21
    ejercicio
    1.14
    =$?
    1.12
     maïs
    1.10
     uſe
    1.09
    problema
    1.08
    regarding
    1.07
     peuvent
    1.07
    Simulated
    1.04
     dessous
    1.03
    Activations Density 0.049%

    No Known Activations