Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 68662
    Prev
    Next
    INDEX
    Explanations

    words related to mint and other mint-like flavors or characteristics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    mint and related aromatics

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    verwijspagina
    -0.71
     otomatig
    -0.64
    pahan
    -0.61
     مشين
    -0.60
    
    -0.59
    VolleyError
    -0.57
    yyj
    -0.57
     للاسماء
    -0.57
     stockbild
    -0.56
     pagină
    -0.54
    POSITIVE LOGITS
    Mint
    1.09
     Mint
    1.04
     mint
    1.02
    mint
    0.95
     MINT
    0.86
     menthe
    0.65
     peppermint
    0.57
     Peppermint
    0.53
     minted
    0.53
    amint
    0.50
    Activations Density 0.002%

    No Known Activations