Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 14364
    Prev
    Next
    INDEX
    Explanations

    references to surveillance and monitoring systems

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

    associated with "control"

    np_acts-logits-general · gemini-2.0-flash

    control of systems and places

    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
     дописавши
    -0.64
     propOrder
    -0.63
    istoitu
    -0.54
    NameInMap
    -0.53
    addCriterion
    -0.53
    newArrayList
    -0.53
     AssemblyProduct
    -0.52
    webElementXpaths
    -0.52
    TintMode
    -0.51
    Broke
    -0.49
    POSITIVE LOGITS
     control
    1.40
     Control
    1.29
    control
    1.20
    Control
    1.14
     CONTROL
    1.13
    控制
    1.07
     controller
    1.04
     command
    1.02
    CONTROL
    1.02
     controllers
    1.02
    Activations Density 0.307%

    No Known Activations