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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 14693
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate simplicity or ease of use

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

    "easy" or its variations

    np_acts-logits-general · gemini-2.0-flash

    easy to describe actions

    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
     ProtoMessage
    -0.74
    randomUUID
    -0.67
    contentLoaded
    -0.66
     hObject
    -0.66
    ագրություններ
    -0.65
    WireFormatLite
    -0.63
     AssemblyTitle
    -0.63
     kasarigan
    -0.62
    chromedriver
    -0.61
     تانيه
    -0.59
    POSITIVE LOGITS
    EASY
    1.14
     Easy
    1.12
    Easy
    1.08
     easy
    1.03
     EASY
    1.03
    easy
    0.99
     access
    0.93
     Easily
    0.92
     Ease
    0.92
     Eas
    0.91
    Activations Density 0.101%

    No Known Activations