© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    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. 28208
    Prev
    Next
    INDEX
    Explanations

    references to special or particular characteristics and attention

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

    special and particular things

    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
    SourceChecksum
    -0.46
    蚪
    -0.44
    bibfield
    -0.43
    )]{
    -0.42
    Fazit
    -0.42
     propOrder
    -0.42
    Install
    -0.42
     completa
    -0.41
    protoimpl
    -0.40
    FormTagHelper
    -0.39
    POSITIVE LOGITS
     special
    1.03
     especial
    0.90
     especiales
    0.80
    special
    0.80
     specialized
    0.78
    特別な
    0.77
     spécial
    0.76
     spéciaux
    0.75
     particular
    0.74
     specific
    0.72
    Activations Density 0.039%

    No Known Activations