© 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. 287
    Prev
    Next
    INDEX
    Explanations

    programming language names

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

    The neuron activates on uppercase identifiers and proper names for technologies—e.g. programming‐language names, tools, modules, or scenario labels.

    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
     or
    -1.64
     first
    -1.48
     schlä
    -1.36
     which
    -1.30
     from
    -1.29
     all
    -1.29
     their
    -1.28
     Remember
    -1.27
     at
    -1.25
     wholly
    -1.23
    POSITIVE LOGITS
    готовка
    1.78
    igth
    1.53
    SOUNDBITE
    1.53
    皛
    1.50
     iria
    1.41
    
    1.39
    籏
    1.38
    蹰
    1.35
    ですが
    1.34
     ´
    1.32
    Activations Density 0.000%

    No Known Activations