© 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. 118272
    Prev
    Next
    INDEX
    Explanations

    code and configuration files

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

    The neuron fires primarily on technical “long” tokens such as URLs, file‐or package‐paths, import statements, or other code/config identifiers (often containing digits or punctuation).

    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
     what
    -0.86
     trying
    -0.84
    windig
    -0.82
     took
    -0.80
     facilitating
    -0.79
     collaborative
    -0.79
     creates
    -0.77
     creating
    -0.77
     ramah
    -0.77
     facilitate
    -0.77
    POSITIVE LOGITS
    ouc
    0.88
    alist
    0.84
     terv
    0.83
     nasce
    0.81
     PORTRAIT
    0.79
    吲
    0.79
    ibrill
    0.77
     设计
    0.77
    FTFY
    0.76
     potrebbe
    0.76
    Activations Density 0.001%

    No Known Activations