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-2-2B
    4. Residual Stream - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 944
    Prev
    Next
    INDEX
    Explanations

    function calls or definitions in code

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

    function definitions in programming code

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

    code generation and programming constructs

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
    ing
    -0.97
    ING
    -0.73
    端
    -0.65
    ة
    -0.63
     Teil
    -0.63
    ingl
    -0.61
    ส์
    -0.60
    boz
    -0.60
    bootstrapcdn
    -0.58
    ning
    -0.58
    POSITIVE LOGITS
    setVerticalGroup
    1.05
     InputDecoration
    0.90
    SequentialGroup
    0.89
    verwijspagina
    0.79
    SpringBootTest
    0.76
     createState
    0.73
    出版年
    0.73
     jsPsych
    0.73
     viewDidLoad
    0.72
    />";
    0.70
    Activations Density 0.010%

    No Known Activations