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. 30568
    Prev
    Next
    INDEX
    Explanations

    events that cause delays or disruptions in timing

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

    events delayed or canceled

    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
     ordres
    -0.43
     pinulongan
    -0.43
     geprek
    -0.42
     académico
    -0.42
     embaraz
    -0.42
     pouvoit
    -0.41
    recated
    -0.40
     empleos
    -0.40
     étoit
    -0.40
     variaciones
    -0.40
    POSITIVE LOGITS
    awaiter
    0.50
     noon
    0.48
     weather
    0.47
     noqa
    0.46
     midday
    0.46
     afternoon
    0.45
    MethodManager
    0.44
     WebDriverWait
    0.44
     يتيمه
    0.43
    下午
    0.43
    Activations Density 0.006%

    No Known Activations