© 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. 9171
    Prev
    Next
    INDEX
    Explanations

    Hag followed by noun/medical suffixes

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

    the text string "Hag" followed by different endings like "rid", "ai", "ley" or "ans", often detecting character names or surnames.

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

    The neuron responds to the character sequence “hag” wherever it appears in words or names.

    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
    笵
    -2.84
    臯
    -2.69
    -2.45
    鉨
    -2.34
    熥
    -2.28
    那個
    -2.25
    
    -2.22
    媄
    -2.19
    夕方
    -2.16
    𝅥
    -2.16
    POSITIVE LOGITS
    {
    3.02
    “
    2.36
    鈈
    2.33
    栳
    2.33
     themſelves
    2.30
    a
    2.23
    g
    2.16
    𝐒
    2.16
     itſelf
    2.16
     koning
    2.16
    Activations Density 0.005%

    No Known Activations