© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 93178
    Prev
    Next
    INDEX
    Explanations

    instances of the word "jug" and references to multitasking

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

    multitasking and jugular vein

    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
    unhofer
    -0.37
     secas
    -0.33
     wst
    -0.32
    QRST
    -0.32
     crema
    -0.31
    enoord
    -0.31
     blive
    -0.30
    -0.30
    dostęp
    -0.30
     overeen
    -0.30
    POSITIVE LOGITS
     juggling
    1.19
    Jug
    0.94
     jug
    0.91
     Jug
    0.88
    jug
    0.85
     multitasking
    0.78
     Balancing
    0.71
     combineReducers
    0.70
     balancing
    0.69
    balancing
    0.68
    Activations Density 0.005%

    No Known Activations