© 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. 10-GEMMASCOPE-RES-131K
    7. 113075
    Prev
    Next
    INDEX
    Explanations

    Persian language and places

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

    The neuron selectively activates on the word “Persian,” flagging mentions of anything explicitly labeled Persian.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    儅
    -3.20
    ſtra
    -3.00
     welchem
    -2.88
    山的
    -2.69
    
    -2.59
     juſ
    -2.58
    揠
    -2.50
    咈
    -2.48
    嘬
    -2.48
     fermer
    -2.48
    POSITIVE LOGITS
    a
    3.23
    }$
    2.30
    ة
    2.28
    ”(
    2.27
    ついに
    2.20
     includes
    2.20
        
    2.17
    』
    2.17
    e
    2.16
     $^{\
    2.13
    Activations Density 0.010%

    No Known Activations