© 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. 71151
    Prev
    Next
    INDEX
    Explanations

    arachidonic acid

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

    the letters "id" in words, particularly at the end of word segments.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    This neuron detects the character sequence “id” within tokens.

    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
    er
    -2.89
    a
    -2.77
    al
    -2.75
    呖
    -2.47
    B
    -2.47
    ↵↵
    -2.39
    oquia
    -2.27
    U
    -2.23
    了
    -2.17
    es
    -2.14
    POSITIVE LOGITS
    葑
    2.44
    id
    2.31
    ⫹
    2.30
    艹
    2.28
    </strong>
    2.28
    
    2.28
    的
    2.23
     Zwecke
    2.20
    好吗
    2.19
    sandalia
    2.13
    Activations Density 0.040%

    No Known Activations