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-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 15549
    Prev
    Next
    INDEX
    Explanations

    specific mathematical expressions and variables

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

    Mathematical/statistical formula components

    np_acts-logits-general · gemini-2.0-flash

    mathematical 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-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 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
     journal
    -1.08
    Journal
    -1.04
     justice
    -1.02
     judgment
    -1.00
     joint
    -1.00
    journal
    -0.99
     Journal
    -0.97
     JOURNAL
    -0.96
     judge
    -0.95
     judgement
    -0.92
    POSITIVE LOGITS
    UnifiedTopology
    0.60
    BeginContext
    0.57
    PreferredItem
    0.56
    basicConfig
    0.54
    FormBuilder
    0.54
    aarrggbb
    0.53
     surla
    0.53
    
    
    0.52
    rowsiness
    0.51
    ́ng
    0.51
    Activations Density 0.223%

    No Known Activations