© 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. 34-GEMMASCOPE-RES-131K
    7. 98711
    Prev
    Next
    INDEX
    Explanations

    codes

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

    The neuron is keyed to numeric tokens, i.e. it activates wherever standalone digits or multi‐digit numbers appear.

    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
     October
    -0.88
    较大
    -0.82
    蛮
    -0.81
    rière
    -0.79
    🩴
    -0.79
     Because
    -0.75
     Remember
    -0.75
    almaz
    -0.74
    鏖
    -0.74
     milder
    -0.73
    POSITIVE LOGITS
     codes
    1.98
     code
    1.77
     códigos
    1.57
     코드
    1.54
     Codes
    1.49
     Code
    1.48
    代码
    1.48
    codes
    1.43
    Codes
    1.38
     código
    1.37
    Activations Density 0.020%

    No Known Activations