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-2-9B
    4. Residual Stream - 16k
    5. 21-GEMMASCOPE-RES-16K
    6. 6756
    Prev
    Next
    INDEX
    Explanations

    instances of watching or viewing activities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.hook_resid_post
    Hook Layer
    21
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     meille
    -0.50
    ########.
    -0.49
    brengen
    -0.48
     gyhoeddwyd
    -0.47
     materialen
    -0.42
    zelfde
    -0.40
     Gari
    -0.40
     gelijk
    -0.40
    égalité
    -0.40
     grond
    -0.40
    POSITIVE LOGITS
     Watch
    1.14
     WATCH
    1.11
     watch
    1.06
     watching
    1.06
    Watch
    1.02
    watch
    1.00
     Watching
    0.96
    Watching
    0.95
    watching
    0.95
     Watched
    0.94
    Activations Density 0.128%

    No Known Activations