© 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. 36251
    Prev
    Next
    INDEX
    Explanations

    if statements

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

    The neuron spots programming‐language tokens—keywords, function or constant names, and other identifiers—within code snippets.

    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
     kroppen
    -1.13
     pistes
    -1.06
    opardy
    -1.06
    úť
    -1.03
    ->___
    -1.03
     ſuch
    -1.03
    ць
    -1.03
    ლ
    -1.01
     Ndi
    -1.01
     sime
    -1.01
    POSITIVE LOGITS
     if
    3.52
     If
    2.02
    if
    2.00
     если
    1.61
     whether
    1.60
     यदि
    1.38
     decide
    1.38
     determine
    1.32
     اگر
    1.30
     jeśli
    1.29
    Activations Density 0.008%

    No Known Activations