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. 24-GEMMASCOPE-RES-16K
    6. 8486
    Prev
    Next
    INDEX
    Explanations

    references to documents, records, and legal proceedings

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

    preceding the type/name of something

    np_acts-logits-general · gemini-2.0-flash

    record document, programming languages, presidential campaigns, sister channel

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.24.hook_resid_post
    Hook Layer
    24
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     bienfaits
    -0.80
     dieux
    -0.80
     épaules
    -0.73
     autorités
    -0.70
     vœux
    -0.70
     lèvres
    -0.69
     modalités
    -0.69
     enjeux
    -0.68
     oiseaux
    -0.66
     négociations
    -0.66
    POSITIVE LOGITS
     histories
    0.92
     budgets
    0.89
     schemes
    0.87
     statuses
    0.87
     programs
    0.86
     formats
    0.86
     assemblies
    0.85
     setups
    0.85
     architectures
    0.84
     versions
    0.84
    Activations Density 2.248%

    No Known Activations