© 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-2-9B
    4. Residual Stream - 16k
    5. 2-GEMMASCOPE-RES-16K
    6. 4026
    Prev
    Next
    INDEX
    Explanations

    instances of the term "works" in various contexts, usually in reference to creative or artistic products

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 2-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    FromNib
    -0.56
     '\\;'
    -0.51
    IntoConstraints
    -0.49
    SerializedName
    -0.47
    Попис
    -0.47
     ensue
    -0.46
     GenerationType
    -0.44
     dAtA
    -0.43
     ramach
    -0.43
     flask
    -0.42
    POSITIVE LOGITS
     WORKS
    0.73
     Works
    0.73
    works
    0.71
     works
    0.70
    Works
    0.69
    WORKS
    0.60
     holds
    0.58
     hear
    0.57
    hold
    0.56
     hold
    0.54
    Activations Density 0.112%

    No Known Activations