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

    Neuronpedia

    APICircuit 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. 20-GEMMASCOPE-RES-16K
    6. 14013
    Prev
    Next
    INDEX
    Explanations

    phrases or terms related to advanced technologies and processes in manufacturing

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
     Normdatei
    -0.74
    Apparently
    -0.65
     Apparently
    -0.63
     فريبيس
    -0.63
    (?)
    -0.61
     quelconque
    -0.61
     (?)
    -0.61
     mouseClicked
    -0.60
     jakieś
    -0.59
    Another
    -0.59
    POSITIVE LOGITS
     unique
    1.28
     innovative
    1.15
     specially
    1.08
     uniquely
    1.07
     powerful
    1.07
     stunning
    1.03
     customizable
    1.03
     sophisticated
    1.02
     exquisite
    1.01
     expertly
    1.01
    Activations Density 0.332%

    No Known Activations