© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 28496
    Prev
    Next
    INDEX
    Explanations

    specific items and categories

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

    This neuron fires on prominent title‐like or heading terms—typically proper nouns and key content words that label sections or named entities.

    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
     consumir
    -1.06
     ingenieros
    -1.06
     either
    -1.05
     künf
    -1.02
    マジで
    -0.98
     acreditar
    -0.95
     öffentliche
    -0.94
     folgte
    -0.94
     cauza
    -0.94
     lucro
    -0.93
    POSITIVE LOGITS
     for
    1.41
    Anything
    1.03
    其他
    1.03
     besonders
    1.00
    その他
    0.99
    更多
    0.96
    各种
    0.93
    НЫЕ
    0.92
    anything
    0.92
     with
    0.91
    Activations Density 0.081%

    No Known Activations