© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 74821
    Prev
    Next
    INDEX
    Explanations

    lan tan

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

    The neuron fires on uncommon subword tokens—especially foreign‐language fragments or alphanumeric/technical pieces (e.g. those containing digits or diacritics).

    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
    e
    -2.13
     abiertos
    -1.99
    }}$
    -1.97
     basado
    -1.95
     один
    -1.93
    呖
    -1.88
     chiedere
    -1.88
     ia
    -1.83
    ным
    -1.83
    榖
    -1.81
    POSITIVE LOGITS
    翾
    2.34
     sereia
    2.09
     gewähr
    2.09
     increments
    2.06
    柘
    2.02
    黽
    1.99
     boneca
    1.97
    蘩
    1.96
    
    1.95
    ܖ
    1.90
    Activations Density 0.001%

    No Known Activations