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. 16-GEMMASCOPE-RES-65K
    6. 53119
    Prev
    Next
    INDEX
    Explanations

    complex mathematical notations and formulas

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

    mathematical formulas and symbols

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     MainAxisSize
    -0.75
    rungsseite
    -0.71
    AnchorStyles
    -0.65
    })^{
    -0.63
    })^
    -0.61
    NOPQRST
    -0.60
    itarias
    -0.60
    '])){
    
    -0.59
    ]^
    -0.58
     endforeach
    -0.58
    POSITIVE LOGITS
    crollView
    0.51
    isfile
    0.44
    ustering
    0.44
     hielt
    0.42
    Опис
    0.41
    Upgrade
    0.40
    可以通过
    0.40
    ne
    0.40
    Up
    0.40
    JsonNode
    0.40
    Activations Density 0.808%

    No Known Activations