© 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 - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 11125
    Prev
    Next
    INDEX
    Explanations

    expressions related to health and medical terminology

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

    Preceding nouns, adjectives, or numbers

    np_acts-logits-general · gemini-2.0-flash

    descriptive adjective-noun pairs

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 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
    읖
    -0.72
    matchCondition
    -0.71
    ðsíða
    -0.70
    ſelben
    -0.69
     müſſen
    -0.68
    ſchaft
    -0.68
    citroën
    -0.68
    WriteTagHelper
    -0.68
    खरीदारी
    -0.67
    principalColumn
    -0.67
    POSITIVE LOGITS
     of
    0.27
     hotel
    0.26
     métiers
    0.26
    ,
    0.23
     past
    0.23
     OF
    0.23
     sorts
    0.23
     stanovnika
    0.23
     infection
    0.22
    FormState
    0.22
    Activations Density 0.761%

    No Known Activations