© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 28-GEMMASCOPE-RES-16K
    11. 4409
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate the effort or time required for tasks or actions

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

    following "take" or "took"

    np_acts-logits-general · gemini-2.0-flash

    takes time and effort

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.28.hook_resid_post
    Hook Layer
    28
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     hinaus
    -0.54
     houſe
    -0.51
    ArrowToggle
    -0.49
     externi
    -0.44
    PostExecute
    -0.43
    RTEE
    -0.42
     ")");
    -0.42
     culturel
    -0.41
    tawesome
    -0.41
    RegistryLite
    -0.41
    POSITIVE LOGITS
     awhile
    0.50
     effort
    0.50
    takes
    0.50
     years
    0.48
     YEARS
    0.45
     takes
    0.44
    years
    0.41
     weeks
    0.41
     jaren
    0.40
    Takes
    0.40
    Activations Density 0.076%

    No Known Activations