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. 20-GEMMASCOPE-RES-16K
    6. 11865
    Prev
    Next
    INDEX
    Explanations

    variations of the word "relate" or its derivatives

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

    Starts with "Rel" followed by other letters

    np_acts-logits-general · gemini-2.0-flash

    relator and other rel- words

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    scalatest
    -0.61
    のは
    -0.59
    BURGH
    -0.59
    autoconfigure
    -0.55
     actionPerformed
    -0.54
    enumi
    -0.53
    cách
    -0.53
     immediate
    -0.52
    هر
    -0.52
    DotNetBar
    -0.51
    POSITIVE LOGITS
     rel
    0.96
     REL
    0.89
     Rel
    0.83
    REL
    0.77
    rel
    0.75
    DeleteBehavior
    0.75
     relati
    0.72
    Rela
    0.69
    Rel
    0.65
     Rela
    0.64
    Activations Density 0.084%

    No Known Activations