© 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. 8301
    Prev
    Next
    INDEX
    Explanations

    code syntax structures and operators commonly used in programming and data processing

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

    Code diff indicators

    np_acts-logits-general · gemini-2.0-flash

    code diffs

    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
    
    -0.62
     Pingback
    -0.59
     ProtoMessage
    -0.59
    tableFuture
    -0.58
    aarrggbb
    -0.58
     CreateTagHelper
    -0.57
    RunAsync
    -0.57
    CRAFT
    -0.57
    cookieParser
    -0.57
    kháu
    -0.56
    POSITIVE LOGITS
    المكان
    0.49
     Waray
    0.47
    kloped
    0.44
    ScopeManager
    0.42
    علق
    0.41
     tué
    0.41
    Đi
    0.40
     pelota
    0.40
     eût
    0.39
     åt
    0.39
    Activations Density 0.123%

    No Known Activations