© 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. 34-GEMMASCOPE-RES-131K
    7. 99117
    Prev
    Next
    INDEX
    Explanations

    lost control

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

    The neuron activates on numeric tokens, especially multi‐digit numbers.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     Yates
    -0.76
    @@@@@@@@
    -0.75
    wani
    -0.75
    Espa
    -0.70
    favor
    -0.69
    verlust
    -0.69
    formal
    -0.67
    Spaghetti
    -0.67
     Kish
    -0.66
    忽略
    -0.66
    POSITIVE LOGITS
     Returning
    0.72
    返回
    0.69
     Citations
    0.68
     返回
    0.68
    𝙐
    0.67
    Returning
    0.65
     returning
    0.62
    犬
    0.62
     kHz
    0.61
    0.60
    Activations Density 0.047%

    No Known Activations