© 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. 4-GEMMASCOPE-RES-16K
    6. 524
    Prev
    Next
    INDEX
    Explanations

    adverbs that describe careful or precise actions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.4.hook_resid_post
    Hook Layer
    4
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     hende
    -0.61
     optimisation
    -0.60
     Monuments
    -0.57
     Galleria
    -0.55
     weathering
    -0.55
    есть
    -0.55
     nodding
    -0.54
     brightened
    -0.54
     Comparable
    -0.53
     Zä
    -0.53
    POSITIVE LOGITS
    correctly
    0.86
     safely
    0.83
    cibly
    0.81
     directly
    0.77
     successfully
    0.75
     silently
    0.74
    )._
    0.74
    conditionally
    0.72
    directly
    0.72
     efficiently
    0.72
    Activations Density 0.496%

    No Known Activations