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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 8194
    Prev
    Next
    INDEX
    Explanations

    syntactical structures associated with programming functions and conditions

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

    Code snippets and code-related terms

    np_acts-logits-general · gemini-2.0-flash

    end of statement; or period

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
     autorytatywna
    -0.98
    تقاوى
    -0.70
    ThroughAttribute
    -0.67
    DIPSETTING
    -0.65
    enumii
    -0.63
     vVar
    -0.61
     bVar
    -0.61
    évaluateur
    -0.60
    sidemargin
    -0.59
    ंदीखरीदारी
    -0.58
    POSITIVE LOGITS
     either
    0.36
     amada
    0.33
    ContentAsync
    0.32
    either
    0.31
     entweder
    0.31
    soup
    0.29
     ));
    0.28
     fumée
    0.28
     broccoli
    0.28
      
    0.27
    Activations Density 0.403%

    No Known Activations