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. 35435
    Prev
    Next
    INDEX
    Explanations

    instances of engagement with complex cognitive processes

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

    keywords for problem solving

    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
    ::_('
    -0.32
    ily
    -0.31
    望
    -0.28
    port
    -0.28
    ρέπει
    -0.26
    nativeElement
    -0.25
    其
    -0.25
     Bal
    -0.25
     вар
    -0.24
    észetes
    -0.24
    POSITIVE LOGITS
     EconPapers
    0.78
     HasFactory
    0.67
    webElementXpaths
    0.64
     ſein
    0.60
     ſehr
    0.59
    WebElementEntity
    0.58
     ſei
    0.57
     <<<<<<<<<<<<<<
    0.57
    ſchen
    0.57
     feroit
    0.56
    Activations Density 0.133%

    No Known Activations