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. 21-GEMMASCOPE-RES-16K
    6. 2108
    Prev
    Next
    INDEX
    Explanations

    phrases indicating time, specifically references to the "last" or "this" in relation to events or periods

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70
    Prompts (Dashboard)
    36,864 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
    Conclusie
    -0.57
     autorytatywna
    -0.56
    jooq
    -0.54
    exitRule
    -0.52
    addCriterion
    -0.50
     AssemblyProduct
    -0.49
    లాలు
    -0.49
     CreateTagHelper
    -0.48
    Fazit
    -0.46
     iprot
    -0.45
    POSITIVE LOGITS
     week
    2.68
     year
    2.44
     month
    2.42
     weekend
    2.01
    week
    1.98
     morning
    1.84
     semaine
    1.81
    month
    1.80
     evening
    1.79
     Woche
    1.76
    Activations Density 0.304%

    No Known Activations