© 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. 1634
    Prev
    Next
    INDEX
    Explanations

    words and phrases related to programming concepts and data structures

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

    programming and technical terms

    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
    Còn
    -0.28
    FieldDescriptor
    -0.28
    RTEX
    -0.27
     suspen
    -0.26
    MessageOf
    -0.26
    WriteAttribute
    -0.26
    He
    -0.25
    wię
    -0.24
     Савезне
    -0.22
    Men
    -0.22
    POSITIVE LOGITS
    ſelben
    0.77
     パンチラ
    0.74
     メンテナ
    0.74
     ſeines
    0.73
     Dieſe
    0.73
    
    0.73
    ꪼ
    0.73
     ſeyn
    0.73
    <unused11>
    0.72
    <unused14>
    0.72
    Activations Density 0.208%

    No Known Activations