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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 85651
    Prev
    Next
    INDEX
    Explanations

    references to the European Commission and related EU entities

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

    European organizations and references

    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
    ̈́
    -0.46
     unknownFields
    -0.42
     propOrder
    -0.37
     artístico
    -0.36
    DECREF
    -0.35
    StructEnd
    -0.35
    vrigt
    -0.34
     Espagne
    -0.34
     oğ
    -0.34
     étoit
    -0.33
    POSITIVE LOGITS
     Administrativna
    0.89
     sectoral
    0.63
     European
    0.63
    äische
    0.62
     trade
    0.60
     Universal
    0.60
     sector
    0.59
     Grand
    0.59
     Federal
    0.59
     federal
    0.58
    Activations Density 0.004%

    No Known Activations