© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 72967
    Prev
    Next
    INDEX
    Explanations

    South Africa

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

    The neuron is triggered by tokens that look like identifiers or names—i.e. any token containing digits or uppercase letters (proper nouns, acronyms, file names, or numbers).

    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
     representan
    -0.88
     consideran
    -0.85
     integran
    -0.82
     between
    -0.82
     chanh
    -0.80
     above
    -0.80
     τῶν
    -0.79
    Dette
    -0.79
    ترف
    -0.78
     secondary
    -0.76
    POSITIVE LOGITS
    preise
    0.86
    kologi
    0.85
    觊
    0.84
     tegens
    0.84
    uteur
    0.82
    IDAD
    0.82
    蜕
    0.82
    êté
    0.81
    ❛
    0.81
    恶心
    0.80
    Activations Density 0.288%

    No Known Activations