Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 65k
    5. 17-GEMMASCOPE-RES-65K
    6. 28546
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions involving square roots and arithmetic operations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.17.hook_resid_post
    Hook Layer
    17
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    awtextra
    -0.50
     Económica
    -0.44
    lieben
    -0.44
    Rhestr
    -0.43
    üman
    -0.42
    </h5>
    -0.41
    setLevel
    -0.41
    իվ
    -0.41
     quirúrg
    -0.40
    RegistryLite
    -0.40
    POSITIVE LOGITS
     bezeichneter
    0.85
    abestanden
    0.71
    SEGUIR
    0.70
    principalTable
    0.67
    RegressionTest
    0.67
    }\]
    0.64
    latego
    0.63
     Paglinawan
    0.63
    NUMX
    0.62
     Réponses
    0.61
    Activations Density 0.215%

    No Known Activations