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-2-9B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 389
    Prev
    Next
    INDEX
    Explanations

    code structure and object properties

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

    identifiers and variable names in programming code, especially those with string or numeric types.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    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
    enterOuterAlt
    -0.46
    OneToOne
    -0.44
    ChildScrollView
    -0.44
     autorytatywna
    -0.40
    Linki
    -0.40
    chyma
    -0.40
     HttpNotFound
    -0.39
     Wikispecies
    -0.39
     okuyayım
    -0.38
    Przypisy
    -0.36
    POSITIVE LOGITS
     kasarigan
    0.78
    toFixed
    0.48
     localisation
    0.43
    CharField
    0.43
    charCodeAt
    0.43
    UniformLocation
    0.42
    conio
    0.41
    jsii
    0.41
    EDEFAULT
    0.40
    lwjgl
    0.40
    Activations Density 0.429%

    No Known Activations