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. 68233
    Prev
    Next
    INDEX
    Explanations

    technical terms related to flexible materials and machinery components

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

    snap-action switch, ceramic material, glass ionomer, pocket diapers, cysteine protease, ion trap, clutch technology, liquid ring pump, radial arm saw, PF resins, blue fans

    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
     Roskov
    -0.58
     كومونز
    -0.57
    Personendaten
    -0.53
    ApiModelProperty
    -0.50
    LookAnd
    -0.50
    reportWebVitals
    -0.48
     createStore
    -0.45
    WebElementEntity
    -0.45
    WriteTagHelper
    -0.45
    ssent
    -0.45
    POSITIVE LOGITS
    LLocation
    0.34
    qtt
    0.32
     hierba
    0.31
    anglès
    0.31
     readline
    0.31
    yczą
    0.31
     мәкалә
    0.30
     inertial
    0.29
    rungsseite
    0.29
    transQ
    0.28
    Activations Density 0.031%

    No Known Activations