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. 3653
    Prev
    Next
    INDEX
    Explanations

    technical terms related to sensors and their functionalities

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

    preceding sensor types or related terms

    np_acts-logits-general · gemini-2.0-flash

    barometer and accelerometer sensors

    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
     оригіналу
    -0.65
     >=",
    -0.64
    ItemBackground
    -0.64
    UserScript
    -0.64
     ModelExpression
    -0.64
    脚注の使い方
    -0.63
    ftagPool
    -0.63
     webpack
    -0.62
    DispatchToProps
    -0.61
     fundido
    -0.60
    POSITIVE LOGITS
     barometer
    0.89
     sensor
    0.89
     sensors
    0.83
     Sensor
    0.81
    Sensor
    0.77
     readings
    0.75
     Sensors
    0.71
    sensor
    0.70
     bar
    0.69
     sensores
    0.69
    Activations Density 0.388%

    No Known Activations