© 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. 34-GEMMASCOPE-RES-131K
    4. 38472
    Prev
    Next
    INDEX
    Explanations

    toggle. toggle

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

    The neuron is specialized to spot code‐style assignments and related operators (e.g. “=”, variable names, bitwise/comparison syntax) in programming text.

    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
     your
    -0.93
     בז
    -0.81
     these
    -0.77
     still
    -0.73
    zij
    -0.71
    toContain
    -0.71
     Have
    -0.69
     Maybe
    -0.69
     exceed
    -0.69
    WithEmail
    -0.68
    POSITIVE LOGITS
     opposite
    1.80
    opposite
    1.48
     ^=
    1.47
    toggle
    1.41
    =!
    1.38
     !
    1.34
     противополо
    1.30
     not
    1.30
    Opposite
    1.26
    !
    1.19
    Activations Density 0.008%

    No Known Activations