© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 92972
    Prev
    Next
    INDEX
    Explanations

    mathematical formulas and code

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

    This neuron detects code‐specific symbols and tokens—things like programming operators (e.g. exponentiation), regular‐expression delimiters, or other non-natural-language syntax.

    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
     चाहिए
    -1.01
     most
    -1.01
    뿐
    -0.98
     since
    -0.98
    諜
    -0.97
     latest
    -0.93
     скоро
    -0.92
     just
    -0.92
     senaste
    -0.91
    esModule
    -0.91
    POSITIVE LOGITS
     homicidio
    1.20
    ,—
    1.05
    為に
    1.01
     والتي
    1.00
    様な
    1.00
    {{{
    0.99
     มัน
    0.97
    θεί
    0.96
    ampiyon
    0.96
    (
    
    0.95
    Activations Density 0.000%

    No Known Activations