© 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. 13589
    Prev
    Next
    INDEX
    Explanations

    numbers and symbols

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

    This neuron responds to numeric tokens—especially multi‐digit numbers (e.g., coefficients, computed numeric answers, and units like “GHz”).

    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
    O
    -1.23
    ]};
    -1.20
    “
    -1.14
    W
    -1.13
    数量
    -1.13
    N
    -1.13
    Our
    -1.12
    Bibliograf
    -1.11
    on
    -1.09
    is
    -1.09
    POSITIVE LOGITS
    undaki
    1.28
    NOME
    1.28
    ambassade
    1.14
     hôte
    1.13
    姈
    1.11
    TIPO
    1.11
    薔薇
    1.09
    vily
    1.09
    
    1.07
     savu
    1.05
    Activations Density 0.149%

    No Known Activations