© 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. 67058
    Prev
    Next
    INDEX
    Explanations

    detecting events and conditions

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

    The neuron fires whenever it sees the substring “detect” (in any form such as detect, detected, detection, Detect) in code identifiers or prose.

    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
    ↵↵
    -2.84
    岀
    -2.41
     évolution
    -2.27
    ৢ
    -2.27
     然
    -2.08
    ."
    -2.06
     tilting
    -2.05
    subsection
    -2.05
     они
    -2.02
     predominant
    -1.98
    POSITIVE LOGITS
    部分は
    3.36
    部分が
    2.94
    ︲
    2.75
    谖
    2.63
    ciutto
    2.50
     możesz
    2.48
     pessoais
    2.45
    ኻ
    2.41
     permitem
    2.41
    也会
    2.41
    Activations Density 0.024%

    No Known Activations