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-2-2B
    4. Residual Stream - 16k
    5. 7-GEMMASCOPE-RES-16K
    6. 4493
    Prev
    Next
    INDEX
    Explanations

    references to mathematical utilities and functions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.7.hook_resid_post
    Hook Layer
    7
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    (/
    -0.46
    Pros
    -0.46
     Chwiliwch
    -0.46
    loi
    -0.41
     ~
    -0.41
     Er
    -0.41
    gså
    -0.41
     transfieras
    -0.41
    迎接
    -0.41
     kuni
    -0.41
    POSITIVE LOGITS
    ContextCompat
    0.92
     LoggerFactory
    0.89
    ioutil
    0.83
    Collectors
    0.79
     UnityEditor
    0.76
    BitmapFactory
    0.76
    IOUtils
    0.76
    Przypisy
    0.73
    TemporalType
    0.72
    }")]
    0.71
    Activations Density 0.099%

    No Known Activations