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. 15-GEMMASCOPE-RES-16K
    6. 1421
    Prev
    Next
    INDEX
    Explanations

    elements and syntax associated with programming or markup languages

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     all
    -0.57
    burgo
    -0.51
    simpleType
    -0.49
    adze
    -0.48
    descend
    -0.48
     mã
    -0.48
    IFEST
    -0.48
    DotNetBar
    -0.48
    NOWLEDG
    -0.47
     bere
    -0.47
    POSITIVE LOGITS
     Beide
    1.17
     Both
    1.08
    Both
    1.08
    どちらも
    1.05
     entrambi
    1.03
     Ambos
    1.01
    Ambos
    0.99
     begge
    0.96
    both
    0.94
     ambos
    0.94
    Activations Density 0.679%

    No Known Activations