© 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. 10-GEMMASCOPE-RES-131K
    7. 102146
    Prev
    Next
    INDEX
    Explanations

    mathematical equations and code

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

    The neuron is specialized to pick out numeric literals—especially larger (multi-digit) numbers—embedded in formulas or code.

    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
    I
    -2.78
    .
    -2.69
    how
    -2.64
    </h4>
    -2.56
    ！”
    -2.53
     奶油
    -2.41
    },
    -2.41
    ization
    -2.38
    It
    -2.34
    与此同时
    -2.31
    POSITIVE LOGITS
    
    2.47
    苤
    2.47
    犼
    2.33
    ୈ
    2.19
    茑
    2.17
     сегодняшний
    2.16
     drenched
    2.14
    餮
    2.14
    Ɑ
    2.13
     afferma
    2.11
    Activations Density 0.055%

    No Known Activations