© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 8407
    Prev
    Next
    INDEX
    Explanations

    information related to product specifications and details

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

    Dimensions, measurements, or sizes

    np_acts-logits-general · gemini-2.0-flash

    dimensions and specifications

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    RegressionTest
    -0.75
    SharedDtor
    -0.66
    tagext
    -0.65
     betweenstory
    -0.63
    saraba
    -0.61
    AddTagHelper
    -0.61
     rhestr
    -0.61
    ########.
    -0.60
    AllAfrica
    -0.60
     Wikimedijinoj
    -0.59
    POSITIVE LOGITS
     approx
    0.56
    Handle
    0.55
     unfolded
    0.52
     montado
    0.52
    overall
    0.52
     overall
    0.51
     Handle
    0.50
     unas
    0.50
    handle
    0.49
    Item
    0.48
    Activations Density 0.035%

    No Known Activations