© 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. 346
    Prev
    Next
    INDEX
    Explanations

    specific entities and acronyms

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

    The neuron consistently activates on capitalized named entities—proper nouns and acronyms (e.g. project names, organizations, people, places).

    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
     these
    -1.18
     this
    -1.02
     deben
    -0.99
     each
    -0.99
     their
    -0.99
     Different
    -0.98
     should
    -0.96
    ᩠
    -0.95
     YOU
    -0.94
     TWO
    -0.94
    POSITIVE LOGITS
     về
    1.13
    
    1.04
     锅
    1.02
     betreff
    1.00
    
    0.98
    关于
    0.98
     Departement
    0.97
     모든
    0.95
    に関しては
    0.95
     departement
    0.93
    Activations Density 0.091%

    No Known Activations