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. 103
    Prev
    Next
    INDEX
    Explanations

    references to programming elements and properties in data structures

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

    complex technical data like IP addresses, error codes, or mathematical expressions.

    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
    Diweddarwch
    -0.48
    .*")]
    -0.47
     betweenstory
    -0.47
     calendriers
    -0.47
     المعيارى
    -0.46
    enumii
    -0.46
    EndGlobalSection
    -0.44
    PreferredItem
    -0.44
     lenker
    -0.43
    adaptiveStyles
    -0.43
    POSITIVE LOGITS
     holder
    0.50
    STELL
    0.45
    holder
    0.41
    edificio
    0.40
    congrat
    0.39
    Cope
    0.38
    Cop
    0.38
    ony
    0.38
    Clouds
    0.38
    copyTo
    0.37
    Activations Density 0.393%

    No Known Activations