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. 68619
    Prev
    Next
    INDEX
    Explanations

    instances of functions and commands related to saving data in a coding context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
    BagConstraints
    -0.50
    PreferredItem
    -0.46
    gnore
    -0.44
     esternos
    -0.44
     (_)
    -0.39
     ostavi
    -0.39
    }());
    -0.38
    moveToFirst
    -0.37
    eraan
    -0.37
    ]-->
    -0.36
    POSITIVE LOGITS
    AntiForgeryToken
    0.52
    AndEndTag
    0.48
     useAppContext
    0.45
    RegressionTest
    0.44
    ConstraintMaker
    0.43
     sbag
    0.42
    formik
    0.42
    føl
    0.42
    submitted
    0.42
     kaarangay
    0.41
    Activations Density 0.003%

    No Known Activations