© 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. 34-GEMMASCOPE-RES-131K
    7. 64955
    Prev
    Next
    INDEX
    Explanations

    programming code blocks

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

    This neuron fires on structural punctuation in code—especially closing parentheses “))”, braces “{”, the loop‐increment “++)”, and similar syntactic tokens.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     όμως
    -1.10
     however
    -1.02
     but
    -1.00
     azonban
    -0.91
    FirebaseUser
    -0.86
     लेकिन
    -0.82
    However
    -0.81
     però
    -0.81
    isotope
    -0.80
     están
    -0.80
    POSITIVE LOGITS
    aplic
    1.05
     would
    0.99
     then
    0.92
    exao
    0.90
    而且
    0.90
     already
    0.90
    fizierung
    0.89
    LJ
    0.89
     piace
    0.86
     diferente
    0.86
    Activations Density 0.016%

    No Known Activations