© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 44456
    Prev
    Next
    INDEX
    Explanations

    code comments

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

    The neuron activates on standalone numeric tokens—particularly three‐digit numbers—typically found in code comments.

    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
    呖
    -3.13
     یا
    -2.98
     или
    -2.70
    หรือ
    -2.67
    _{
    -2.63
    谖
    -2.61
    辚
    -2.56
     和
    -2.42
    お疲れ様でした
    -2.42
     और
    -2.41
    POSITIVE LOGITS
    ра
    2.28
    
    2.19
     rigue
    2.14
    寕
    2.14
    as
    2.14
    ðsíða
    2.13
    鹛
    2.11
    臏
    2.11
    er
    2.08
     économies
    2.08
    Activations Density 0.005%

    No Known Activations